Home

Rainfall Runoff Library User Guide

1. 2001 2001 2001 2002 2002 2002 2002 Units mm day 1 Format 1 Easting 0 000000 Northing 0 000000 Latitude 0 000000 Longitude 0 000000 Elevation 0 000000 363 0 364 0 365 10 2 10 20 3 00 5 416 Figure 51 Example of Tarsier daily time series format 79 Rainfall Runoff Library User Guide 9 References Bear J 1979 Hydraulics of Groundwater McGraw Hill Inc New York ISBN 0 07 004170 9 Boughton W J 1996 AWBM water balance model calibration and operation manual CRC for Catchment Hydrology Burnash R J C Ferral R L and McGuire R A 1973 A Generalised Streamflow Simulation System Conceptual Modeling for Digital Computers U S Department of Commerce National Weather Service and State of California Department of Water Resources Duan Q 1994 The SCE UA method Journal of Hydrology 158 265 284 Grayson R B Argent R M Nathan R J McMahon T A Mein R G 1996 Hydrological recipes estimation techniques in australian hydrology Cooperative Research Centre for Catchment Hydrology p 78 Hooke R and Jeeves T A 1960 Direct Search Solution of Numerical and Statistical Problems Westinghouse Research Laboratories Pittsburgh Pennsylvannia USA 212 228 Kachroo R K 1992 River flow forecasting Part 5 Applications of a conceptual model Journal of Hydrology 133 141 178 Kachroo R K and Liang G C 1992 River flow forecasting Part 2 Alge
2. 56 rainfall runoff library Issued 3 Nov 03 6 2 Model Calibration How stable is the rating at the site The relationship between height i e how sensitive is a flow estimate to a change in height Looped ratings i e where the rating on the rising and falling limbs of hydrographs are different Sticking gauge i e how believable is it at low flows The highest rated flow After assessing the flow data it may be appropriate to remove unrealistic data from the record prior to calibration It is also important to know what part of the flow range has the least error Calibration and validation period There are two ways of approaching a rainfall runoff model calibration 1 Considering the entire period of record and 2 Considering a proportion of the record for calibration and the other proportion for verification 6 2 1 Calibrating over the entire period The main advantage of doing this is that the most optimum calibration for the available data can be achieved The problems with this is that you have no indication of how the model is likely to perform outside of the calibration period i e how robust is the model For this reason this is not the recommended approach for calibrating a model In some circumstances where there is only a small amount of data available then there may be no other option but to use the entire period for calibration If this is the case then it is important that the robustness of the m
3. H Rainfall in excess of infiltration capacity 1 H9x lt Y Moistare in Soil moisture f soil storage in mm excess Generated capacity 8 surface run off Evaporation Layer 1 rp GXg x n Layer 2 Groundwater Surface run routing off routing component c omponent Linear reservoir Kp Layer 2 25 Total estimated discharge Figure 35 SMAR schematic diagram 4 4 1 Water balance The water balance component uses five parameters to describe the movement of water into and out of a generalised soil column under conditions of atmospheric forcing C Z H Y and T The dimensionless parameter C regulates evaporation from the soil layers Evaporation is assumed to vary as an exponential function of the form C where C lies between O and 1 andi 1 2 3 refers to the successive soil layers That is for a given potential evaporation the first layer can meet that demand at the potential rate the second layer at a rate C the third layer at C etc resulting in a reduction in the soil moisture store in an approximately exponential manner The potential evapotranspiration rate from the top layer conceptually represents evapotranspiration from the interception storage and from the topsoil during periods of negligible capillary resistance The parameter Z mm represents the effective moisture storage capacity of the soil contributing to the run off generation mechanisms E
4. essen eee T EER 27 Figure 24 Observed and simulated data comparison ssssssssse 27 Figure 25 Sensitivity dialogen ier e e et e eh ee Feet ene dd 28 Figure 26 Simulation dia loguearte i 29 Figure 27 File save as window 44 en deeper e rie ee e tada 30 Figure 28 Galibrate graphio eec ott eter e lee ge Nee n 31 Figure 29 Sensitivity gp harrier iee e e dde tet tede Hee aoc o PH e 32 Figure 30 Simulation graphiste etse d e ot redet ea er irte ete EGRE AR e eae ct aes 33 Figure 31 Catchment information dialogus emissie na eee ener 34 Figure 32 Structure of the AWBM rainfall runoff model sse 37 Figure 33 Structure of the Sacramento rainfall runoff model sss 38 Figure 34 Structure of the SIMHYD rainfall runoff model se 39 Figure 35 SMAR schematic diagram ssssssssssssseeeeeeeeeeee e eene N eren nnne 41 Figure 36 Structure of the TANK model sse eee eene nennen 45 Figure 3 Z RosenbroeckfUfiction ous ie ete tret ette epe re bes epe E a ee ee eere teta de 48 Figure 38 Initial base in the parameter space ssssssss ener 49 Figure 39 First orthonormalisation of the base ener 50 TOC TOC Figure 40 Second orthonormalisation of the base sss 50 Figure 41Rainfall and evaporation data scaling sssss 55 Figure 42 Example ofAWBM daily time series format 68 Figure 43 Example of Chiew Siriwadena daily rainfall time
5. AWBM The AWBM is a catchment water balance model that can relate runoff to rainfall with daily or hourly data and calculates losses from rainfall for flood hydrograph modelling Data needs and formats are described in Chapter 1 and Chapter 8 respectively The model uses 3 surface stores to simulate partial areas of runoff The water balance of each surface store is calculated independently of the others Figure 32 The model calculates the moisture balance of each partial area at either daily or hourly time steps At each time step rainfall is added to each of the 3 surface moisture stores and evapotranspiration is subtracted from each store The water balance equation is storen storen rain evap n 1 to 3 1 If the value of moisture in the store becomes negative it is reset to zero If the value of moisture in the store exceeds the capacity of the store the moisture in excess of capacity becomes runoff and the store is reset to the capacity When runoff occurs from any store part of the runoff becomes recharge of the base flow store if there is base flow in the stream flow The fraction of the runoff used to recharge the base flow store is BFI runoff where BFI is the base flow index i e the ratio of base flow to total flow in the stream flow The remainder of the runoff i e 1 0 BFI runoff is surface runoff The base flow store is depleted at the rate of 1 0 K BS where BS is the current moisture in the base flow stor
6. If a spinner is provided by clicking on the up arrow the number will increase and clicking on the down arrow the number will decrease Alternatively a real number can be entered in the field If range checking is activated in the field and the number is outside of the specified bounds then the field will be cleared and set to the closest bound Radio button used to make an exclusive selection from a range of actions Checkbox Used to accept or decline an option When ticked the option is accepted Drop down list box clicking on the y drops down a list box from which any one or several entries can be selected by moving the cursor over the entry and pressing the left mouse button Slider bars The slider bars in the RRL are used to select dates By moving the slider bar the associated date will be updated to the new date Colour bar Clicking on the colour bar will display a colour selection dialogue When a new colour is selected the colour bar and associated parameter colours will be changed to the new colour Directory tree This field is used to browse and select files This field operates in a similar manner to windows explorer 11 Rainfall Runoff Library User Guide Time series data field In the top left hand corner the type Rainfall menu and time span of data is displayed To the left of this is a 11111902 graph of the time series data with time on the x axis and 31121917 times series data units
7. Location Input Null value Parameter Calibration dates Validation dates Calibration type o ON BO OH A C ND Recorders o Data scaling and Comments Model The model section contains a string identificating of the type of model used and its version Location The location section contains information about the catchment The name is purely informative while the area in square kilometres is used for automatic conversion between flow in ML day or cubic meters per seconds to runoff depth in mm day Input The input section stores the filename and relative path to the loaded inputs for the model associated with this project This approach was adopted to account for current practices data formats and to avoid duplication of the same input data Note that this means that moving a project file and or data from one location to another may make those relative paths obsolete NullValue This value is used to identify records in data files that are missing Any records thaqt have this value will be assumed to be missing Parameter The parameter section stores the parameter name value lower bound upper bound and fixed flag Note the parameter list varies dependent upon the model type specified The fixed flag is used to identify whether the model parameter is changed during an optimisation 71 Rainfall Runoff Library User Guide Calibration dates Specifies the minimum and maximum possib
8. Sensitivity dialogue The sensitivity dialogue as shown in Figure 25 allows the user to analyse the sensitivity of a particular model parameter with regard to a selected objective function There are four sections to this area 1 Model parameter selection drop down list box 2 Parameter bounds fields 3 Objective function selection drop down list box and 27 Rainfall Runoff Library User Guide 4 Plot curve response button The sensitivity dialogue has two areas a sensitivity analysis parameter area and a graph window Model Input Dates Calibration Sensitivity Simulation Sensitivity Analysis Clear Plot Select a model parameter to inspect Select Parameter To Inspect Parameter Information Min Calibrated value Max Select an Objective Function Sum of squares of errors Change Objective Parameters Plot Curve Response gt gt Advanced Tools Figure 25 Sensitivity dialogue The model parameter selection drop down list box allows the user to select the model parameter to be analysed Note the parameters displayed in this list are dependent upon the rainfall runoff model selected The parameter information section displays the selected parameter bounds and current value The current bounds can be modified by typing the new bound in the appropriate field By clicking the mouse on the objective function drop down list the user can select from the following objective funct
9. The dynamic update check box allows the user to decide whether the graph is updated each time a calibration parameter is changed or not If checked the graph is updated when a parameter is changed This is an extremely useful facility for investigating how the model behaves as different parameters are changed It also gives the user an idea of how sensitive the model is to changing each of the parameters This can be quite slow if the calibration period is long and the model is complex In such cases the calibration period could be shortened for investigation purposes and then increased when full calibration is required 6 3 2 List of calibration parameters This list will vary depending on which model has been selected The parameter values can be incremented by using the spinner next to the parameter or for fine tuning the parameter can be typed in 59 Rainfall Runoff Library User Guide 6 4 6 3 3 Update graph If the run time of the model is quite long it can take some time for the graph to update if dynamic update has been set If dynamic update has not been set activating this button will run the model with the current parameters and update the graph Automatic calibration of models There are two tabs in the calibration dialogue that allow for automatically calibrating models Generic and Custom The generic tab as the name suggest applies to all models while the Custom facility is a pupose built calibrator for a particular
10. Rosenbrock searches is taken as the global optimum Genetic algorithm The genetic algorithm is a search procedure based on the mechanics of natural selection and natural genetics which combines an artificial survival of the fittest with genetic operators abstracted from nature Holland 1975 The genetic algorithm searches among a population of points and works with a coding of the parameter set rather than the parameter values themselves It uses probabilistic transition rules A population of m 100 points are chosen initially at random in the search space The objective function values are calculated at all points and compared From these points two points are selected at random The selected points are subsequently used to generate a new point in a certain random manner with occasionally added random disturbance This is repeated until m 100 new points are generated The generated points are expected to be concentrated in the vicinity of the optima than the original points The new population points which can again be used to generate another one and so on yielding more and more points in the vicinity of the optima 5 6 1 Parameter coding The genetic algorithm works with parameter coding The method of parameter coding that has been used is called binary coding An l bit binary variable is used to represent one parameter xi The integer of the decoded binary variable ranges from O to 2I 1 and can be mapped linearly to the parameter
11. are valid options for the y axis The axis properties that the user may change include e Background colours 14 rainfall runoff library Issued 3 Nov 03 Using RRL e Border colours e Axis ticks spacing length and colour e Tick labels Type precision font colour e Graticules on off colour and line style There are number of options for specifying background colours that include e Background filled true or false e Background begin colour e Background end colour e Background colour transition angle The background colour may be turned on or off by changing the backwallfilled drop down list field The background is filled by a transition of colour from the begin colour backwallbegincolor to the end colour backwallendcolor The colours are set with the colour setting dialogue as shown in Figure 7 The transition of colours is relative to the lower left corner of the graph The backwallgradientangle specifies the angle in degrees for the transition e g angle 0 is left to right while 270 is bottom to top The border tick font and graticule colours can all be specified with the colour dialogue as show in Figure 7 The tick length on the x and y axis can be changed by clicking the mouse on the appropriate field and typing in a new value 0 0 1 0 The number of ticks on the x and y axis can be changes by clicking the mouse on the appropriate files and typing in a new value The labels for the axis are depen
12. can select from the following objective functions Nash Sutcliffe criterion Coefficient of efficiency Sum of square errors Root mean square error RMSE Root mean square difference about bias Absolute value of bias Sum of square roots Sum of square of the difference of square root Sum of absolute difference of the log Runoff difference in 25 Rainfall Runoff Library User Guide Flow duration curve Base flow method 2 There are three data transformation options e Daily time series no transformation e Monthly volumes daily flows summed over the month e Flow duration Sorted daily values from highest to lowest The custom tab is available for rainfall runoff models that have internal calibration algorithms Currently the only model with this capability is the AWBM model as shown in Figure 22 For all other models this tab is greyed This tab displays a drop down list that allows the user to choose internal calibration method for the AWBM model there is only one option to choose from Generic Custom Manual Custom Calibration Figure 22 Custom tab The manual tab is provided for manual calibration of the selected rainfall runoff model There are three sections to this tab 1 Dynamic update check box 2 Model parameters 3 Update graph button When the Dynamic Update check box is checked any changes in model parameters will cause the rainfall runoff model to be run and the cur
13. characters and day 2 characters format The data value is specified to 3 decimal places 10 characters An example of this format is shown in Figure 43 20011228 0 000 20011229 0 000 68 rainfall runoff library Issued 3 Nov 03 Data storage and file formats 20011230 0 000 20011231 10 200 2002 1 1 0 000 2002 1 2 0 000 2002 1 3 0 000 2002 1 4 16 000 Figure 43 Example of Chiew Siriwadena daily rainfall time series format 8 2 3 Chiew and Siriwadena monthly flow time series format flow An Chiew and Siriwadena monthly flow time series format file is an ASCII text file that contains monthly time series flow data There is no header line in the file The data is organized as date followed by data value The date is in year 4 characters and month 2 characters The data value is specified to 2 decimal places 12 characters An example of this format is shown in Figure 44 197410 46623 00 197411 7366 00 197412 1819 00 1975 1 1168 00 1975 2 1305 00 1975 3 1780 00 Figure 44 Example of Chiew and Siriwadena monthly flow times series format 8 2 4 Comma seperated value CSV daily time series format csv A comma separate value CSV daily time series format file is an ASCII text file that contains daily time series data There is no header line in the file There are two columns of data the first column is a date string in day month year format followed by the time series value The tw
14. depressions wetlands and billabongs and never returns to the river and evaporation from the water surface It also includes the effects of uncertainty in river flow gauging measurements and unaccounted water usage Unregulated rivers All rivers that are not regulated including rivers where the flow is controlled by dams or weirs constructed by urban water suppliers or private users Groundwater Hydrogeology Acronym Term Definition 83 Rainfall Runoff Library User Guide Acronym Term Definition Alluvium Sediments clays sands gravels and other materials deposited by flowing water Streams can make Deposits on riverbeds floodplains and alluvial fans Examples Shepparton Calivil and Renmark Formations in the Lower Murrumbidgee Valley Aquifer Rock or sediment in a formation which is saturated and sufficiently permeable to transmit and release water Examples of major alluvial systems include the Shepparton Calivil and Renmark aquifers in the Lower Murrumbidgee Valley These lie underneath the major irrigation areas of southern NSW Aquifer confined An aquifer that is sandwiched between two layers of impervious materials Aquifer unconfined Also known as water table or phreatic aquifer The water table is the upper boundary of unconfined aquifers Aquitard A low permeability layer that can store groundwater and also transmit it slowly from one aquifer to another Artes
15. difference about bias 5 Absolute value of bias 60 rainfall runoff library Issued 3 Nov 03 Model Calibration 6 Sum of square roots 7 Sum of square of the difference of square root 8 Sum of absolute difference of the log There are four secondary objective function options None 2 Runoff difference in 3 Flow duration curve 4 Base flow method 2 The weightings and parameter values for the secondary objective function can be set by clicking on the parameter button below the drop down menu This will display a functions parameter dialogue where specific objective function parameters can be set including the weighting to be applied to the second objective function The user can also decide to apply the objective function to either daily or monthly time steps If the calibrate on monthly values checkbox is checked the objective function will be evaluated based on the monthly aggregate of the daily data Objective functions Objective functions The equations for each of the objective functions are detailed below The modelled value number i in the time series is noted m i and the observed value is noted by the letter o Nash Sutcliffe The Nash Sutcliffe criterion also known as efficiency is a common mesure of fit in rainfall runoff modelling interpreted as how much better the fit is as if one was using the average of the observed data to represent it The objective function used in the RRL is a slight variat
16. ee inicias 68 8 1 3 Dota sub direcl ry iocis aia as ENSE Use REN 68 A o ON or e mI HM eH den 68 8 1 5 Documentation directory sese 68 8 2 Fileformats co eee etd ii bci adde edet Na t udi e deett bnt 68 8 2 1 AWBM daily time series format sess 68 8 2 2 Chiew and Siriwadena daily rainfall time series format rain 68 8 2 3 Chiew and Siriwadena monthly flow time series format flow 69 8 2 4 Comma seperated value CSV daily time series format csv 69 8 2 5 IQQM daily time series format iqqm eseeeeeesse 69 8 2 6 Rainfall runoff library project file jobf sss 70 8 2 7 QDNR SILO daily time series format txt ssuuuuss 73 8 2 8 XML times series format tsx coooooocccccconocccccnnoncncconononccnonononos 73 8 2 9 SWAT daily rainfall time series format pcp 78 8 2 10 Tarsier daily time series format tts oooccconncnncccnccnnncnoncnnno 78 9 RETErENCE oce aste e oasis em 80 TOC 10 10 1 10 2 10 3 GlOSSORY o M UU 82 Surface water Hydrolodgy e eie ene did 82 Groundwater Hydrogeology esee 83 Sells erate Pe eRe AN 85 TABLE OF FIGURES Figure Ie Example of standard wihdowW sieren aia aAa EAR E epeek rA AEEA EUR e Ode IRR SEO EUN SERES 8 Figure 2 Dropdown menuliskste c tot eoe eb bre tho FOR ent e
17. encoding clr 1 0 SOAP ENV encodingStyle http schemas xmlsoap org soap encoding 73 Rainfall Runoff Library User Guide lt SOAP ENV Body gt lt al TimeSeries id ref 1 xmlns al http schemas microsoft com clr nsassem TIME DataTypes DataTypes 2C 20Version 3D1 0 1392 2858822C220Culture 3Dneutral 2C 20PublicKeyToken 3D5 9b30b0da8a82a84 gt lt state href ref 4 gt lt _name id ref 5 gt precip tts lt _name gt nullValue 9999 nullValue lt minCacheValid gt true lt minCacheValid gt lt minCache gt 0 lt minCache gt lt maxCacheValid gt true lt maxCacheValid gt lt maxCache gt 73 84 lt maxCache gt lt categories href ref 6 gt lt lastAllocatedKey gt 0 lt _lastAllocatedKey gt lt _silence gt false lt _silence gt Data x002B name href ref 5 gt Data x002B nullValue 9999 Data x002B nullValue gt Data x002B minCacheValid gt true lt Data_x002B minCacheValid gt Data x002B minCache gt 0 lt Data_x002B minCache Data x002B maxCacheValid gt true lt Data_x002B maxCacheValid Data x002B maxCache gt 73 84 lt Data_x002B maxCache Data x002B units href ref 7 gt Data x002B categories href ref 6 gt Data x002B categoryKeys href ref 8 gt Data x002B lastAllocatedKey gt 0 lt Data_x002B_ lastAllocatedKey gt Data x002B silence gt false lt Data_x002B_ silence Subject x002B silence gt false lt Subject_x002B_ silence
18. gt SimpleUnit x002B dimensions href ref 38 gt SimpleUnit x002B refUnit href ref 33 gt SimpleUnit x002B scalingFactor gt 1 lt SimpleUnit_x002B_ scalingFactor gt SimpleUnit x002B offset gt 0 lt SimpleUnit_x002B_ offset SimpleUnit x002B label href ref 39 gt SimpleUnit x002B hasCustomLabel gt false lt SimpleUnit_x002B_ hasCustomLabel gt Unit x002B dimensions href ref 16 gt Unit x002B label href ref 39 gt Unit x002B hasCustomLabel gt false lt Unit_x002B_ hasCustomLabel gt a3 RefUnit a3 RefUnit id ref 35 xmlns a3 http schemas microsoft com clr nsassem TIME Core Units TimeCore 2C 20Version 3D1 0 1390 33501 2C 20Culture 3Dneutral 2C 20PublicKeyToken 23D1 7delc0433b5448b lt baseUnit href ref 40 gt dimensions href ref 41 gt refUnit href ref 35 gt lt _scalingFactor gt 1 lt _scalingFactor gt lt _offset gt 0 lt _offset gt lt _label id ref 42 gt s lt _label gt lt _hasCustomLabel gt false lt _hasCustomLabel gt SimpleUnit x002B dimensions href ref 41 gt SimpleUnit x002B refUnit href ref 35 gt SimpleUnit x002B scalingFactor 1 SimpleUnit x002B scalingFactor gt SimpleUnit x002B offset 0 SimpleUnit x002B offset SimpleUnit x002B label href ref 42 gt SimpleUnit x002B hasCustomLabel gt false lt SimpleUnit_x002B_ hasCustomLabel gt Unit x002B dimensions href ref 16 gt Unit x002B label href ref 42 gt Unit x
19. includes optimisation methods and the algorithms used in rainfall runoff models 1 1 2 Structure This User Guide has 10 Chapters which are ordered by the way that a typical user would work through the system Chapters cover 1 Introduction User Guide and RRL Installation Using RRL a description of the menus and dialogues Description of models Description of optimisers Description of optimisers Saving results Data storage and file formats O0 0 NX BO Oo a C N Reference o Glossary Rainfall Runoff Library User Guide 1 2 RRL 1 2 1 Overview The RRL uses daily time series rainfall and evapotranspiration data to generate daily catchment runoff The generator provides several commonly used lumped rainfall runoff models calibration optimisers and display tools to facilitate model calibration 1 2 2 Features The RRL currently contains 6 rainfall runoff models 8 calibration optimisers a choice of 10 objective functions and 3 types of data transformation for comparison against observed data There is a graphical user interface that comprises menus dialogues and graph display tools The rainfall runoff models included in the library are e AWBM e Sacramento Simhyd SMAR SWAT core TANK The calibration optimisers included in the library are e Uniform random sampling e Pattern search Multi start pattern search e Rosenbrock search e Rosenbrock multi start search Genetic algorithm Shuff
20. is important If the model is to be used for flood predictions then choosing a wet period for calibration is important Quite often over long flow records the type of instrument used to record river level will vary In the early records daily read staff gauges or Bristol recorders may have been used It is probably a better option in such cases to calibrate during periods of more accurate flow records and assess how well the model performs during the other periods For example you would be expecting the model to over estimate flows as compared to a daily read staff gauge You may also find a smoother shape in recessions compared to a Bristol recorder There may be significant changes in ratings at a site This can be caused by a change in control at that site The selection of the calibration period should be a period of reasonable stable ratings that best represents the intended period of use for the model Land clearing can have a significant effect on the flow characteristics of a catchment particularly during dry periods The clearing of vegetation generally reduces evapotranspiration rates and increases the proportion of surface runoff The calibration period should be selected based on the land use that is most appropriate for the intended use of the model 58 rainfall runoff library Issued 3 Nov 03 6 3 Model Calibration 6 2 3 Model warm up When rainfall runoff models start some estimate of the contents of each of the soil mo
21. lt al TimeSeries gt lt al TimeSeriesState id ref 4 xmlns al http schemas microsoft com clr nsassem TIME DataTypes DataTypes 2C 20Version 3D1 0 1392 28588 2C 20Culture 3Dneutral 22C 20PublicKeyToken 3D5 9b30b0da8a82a84 gt lt _timeStep href ref 9 gt lt values href ref 10 gt lt _count gt 6209 lt _count gt lt _start gt 0001 01 01T00 00 00 0000000 11 00 lt _start gt lt _end gt 0017 12 31T00 00 00 0000000 11 00 lt _end gt lt al TimeSeriesState gt lt a2 Hashtable id ref 6 xmlns a2 http schemas microsoft com clr ns System Collections gt lt LoadFactor gt 0 72 lt LoadFactor gt lt Version gt 0 lt Version gt lt Comparer xsi null 1 gt lt HashCodeProvider xsi null 1 gt lt HashSize gt 11 lt HashSize gt lt Keys href ref 11 gt lt Values href ref 12 gt lt a2 Hashtable gt lt a3 CompoundUnit id ref 7 xmlns a3 http schemas microsoft com clr nsassem TIME Core Units TimeCore 2C 20Version 3D1 0 1390 33501 2C 20Culture 3Dneutral 2C 20PublicKeyToken 23D1 7de1c0433b5448b gt lt _units href ref 13 gt lt _prefix href ref 14 gt lt _label id ref 15 gt lt _label gt lt _hasCustomLabel gt false lt _hasCustomLabel gt Unit x002B dimensions href ref 16 gt lt Unit x002B label href ref 15 gt Unit x002B hasCustomLabel gt false lt Unit_x002B_ hasCustomLabel gt lt a3 CompoundUnit gt lt a2 Hashtable id ref 8 xmlns a2 http schemas mi
22. parameter function is defined as z 1 xy 100 y x y 1 The global minimum is found at x y 1 and is in a long and narrow valley as illustrated by Figure 37 10D H 1 5 1 0 5 D 0 5 1 1 5 Figure 37 Rosenbrock function 48 rainfall runoff library Issued 3 Nov 03 Description of optimisers Let us start the search at the point x 1 y 0 6 2 The search starts by initialising the base along the coordinate directions of the parameter space as shown by Figure 38 0 8 0 75 0 7 0 65 0 6 0 55 0 5 0 45 0 4 0 35 0 3 1 1 1 0 9 0 8 0 7 0 6 0 5 Figure 38 Initial base in the parameter space The algorithm then searches along these directions Figure 39 shows the successive successful steps in blue lines Each time a step in a direction is successful the step size is increased by a factor gt for the next search in that direction If the attempted step is unsuccessful the step size is multiplied by a factor 1 lt JJ lt 0 Note that is going in the other direction The unsuccessful steps are not shown in the figure but they are taken into account to define the end of the stage A stage is completed once there has been ot least one successful step and unsuccessful step in each direction When the stage is finished the base is rotated via a Gram Schmidt orthonormalisation procedure in order to have one vector collinear
23. rainfall station represents the flow from the catchment is to plot the rainfall and flow on similar scales The rainfall peaks can then be checked against the flow peaks to see if the size of peaks correlates with the amount of rainfall and that the peaks occur at about the same time The number of sites is a very important issue when multiple rainfall sites are available Typically Theissen weightings are used to associate a protion of the catchment with each rainfall station There are a few things to consider More is not always better The more stations you have the more raindays that will occur on the catchment Care should be taken to makes sure that the number of rain days per month is similar to the number of flow peaks per month When generating long flow records it is tempting to have Theissen weightings that vary as rainfall sites cut in and out This approach should be used with extreme caution as calibrating models to different groups of stations generally leads to different calibration parameters Consequently you will have no idea how robust the model is when the number of rainfall sites is considerably different to the period when the model was calibrated Be very careful when using rainfall surface data for the reasons mentioned above It is not recommended that rainfall be used for every grid square that is available in the catchment A better approach is to use the monthly surfaces at each grid square to estimate the ave
24. ref 26 xmlns a2 http schemas microsoft com clr ns System Collections gt lt _items href ref 28 gt lt size 1 size lt version 1 version lt a2 ArrayList gt lt SOAP ENC Array id ref 27 SOAP ENC arrayType xsd anyType 16 gt lt item href ref 29 gt lt SOAP ENC Array gt lt SOAP ENC Array id ref 28 SOAP ENC arrayType xsd anyType 16 gt lt item href ref 30 gt lt SOAP ENC Array gt a3 CompoundUnit x002B PoweredUnit id ref 29 xmlns a3 2 http schemas microsoft com clr nsassem TIME Core Units TimeCore 2C 20Version 3D1 0 1390 33501 2C 20Culture 3Dneutral 2C 20PublicKeyToken 23D1 7delc0433b5448b gt lt _u href ref 31 gt lt _p gt 1 lt _p gt a3 CompoundUnit x002B PoweredUnit gt a3 CompoundUnit x002B PoweredUnit id ref 30 xmlns a3 http schemas microsoft com clr nsassem TIME Core Units TimeCore 2C 20Version 3D1 0 1390 33501 2C 20Culture 3Dneutral 2C 20PublicKeyToken 23D1 7delc0433b5448b gt u href ref 32 gt p 1 p a3 CompoundUnit x002B PoweredUnit lt a3 SimpleUnit id ref 31 xmlns a3 http schemas microsoft com clr nsassem TIME Core Units TimeCore 2C 20Version 3D1 0 1390 33501 2C 20Culture 3Dneutral 2C 20PublicKeyToken 23D1 7delc0433b5448b gt lt dimensions href ref 16 gt lt _refUnit href ref 33 gt lt _scalingFactor gt 0 001 lt _scalingFactor gt lt _offset gt 0 lt _offset gt lt _ label id ref 34 gt mm
25. series format sssssssssss 69 Figure 44 Example of Chiew and Siriwadena monthly flow times series format sss 69 Figure 45 Example of CSV daily time series format 69 Figure 46 Example of IQQM daily time series format sss 70 Figure 47 Example of rainfall runoff library job file format 73 Figure 48 Example of QDNR SILO daily time series format ssse 73 Figure 49 Example of XML times series format ssssssssssseee ee eere 78 Figure 50 Example of SWAT daily rainfall time series format sss 78 Figure 51 Example of Tarsier daily time series format sssssss 79 VI Using RRL 1 Introduction This Chapter introduces you to the Rainfall Runoff Library RRL and this User Guide lt describes the purpose of this User Guide the structure and content of the User Guide e features of the RRL e the data requirements for using the RRL and where to locate other documentation available on the RRL 1 1 The user guide 1 1 1 Purpose This User Guide describes how you interact with the Rainfall Runoff Library RRL It describes what you can do with the system and how you do it While the RRL can be applied to other catchments this User Guide uses several example catchments to provide examples for setup navigation interrogation calibration and analyses of model results This User Guide also describes how the model works that
26. term used to describe a group of soils that can be managed similarly and which exhibit similar features and properties 85
27. the fluxes out of the model 6 1 2 Rainfall data The calibration of a rainfall runoff model is most sensitive to the rainfall data that is provided If the volume of rainfall is incorrect or the rain days are not representative of the peaks in flow then calibration may be difficult with very poor results There are several things that need to be considered in the preparation of rainfall data Catchment average rainfall Selection of appropriate rainfall sites Catchment average rainfall The catchment average rainfall can be estimated by many different methods two of the more common methods are discussed here The first method is to draw a isoheytal map across the catchment and the second method is sum grid squares from a rainfall surface spline An isoheytal map is basically a contour map of typically average annual rainfall Drawing an isoheytal map is a relatively easy process when there are a number of gauges in and surrounding the catchment Care should be taken to ensure all rainfall sites are gap filled and that the period selected is common to alll sites Climate databases such as SILO have splined surfaces that cover Australia These surfaces take into account location distance from coast and elevation to derive average annual rainfall across grid squares This can be summed for the grid squares in a catchment and averaged Rainfall site selection There are several things that need to be considered in selecting rain
28. the down arrow the number will decrease 10 rainfall runoff library Issued 3 Nov 03 18 July 2003 EH Er Mon Tue Wed Thu Fr Sat Sun 1 2 3 4 5 6 8 9 WO Ail 12 13 14 15 16 E 18 19 20 21 22 23 24 25 26 e 28 29 30 31 January 2003 C Today 18 07 2003 Calibrate on duration curve Pattern Search Multi Start Genetic Algorithm SCE UA e Pattern Search Multi Start Pattern Search L C Program Files Toolkit Rainfall Runot DLC 070025 EVG iqqm 070025 ptfigqm CalculatedFlowsDLNWC tts PET tts Rainfall tts RealFlow_412050 tts Mekong NSW NT QLD calibration Fl RH E Using RRL by one Alternatively an integer can be entered in the field f range checking is activated in the field and the number is outside of the specified bounds then the field will be cleared and set to the closest bound Date fields There are two ways that a date can be entered The date field is divided into three parts separated by day month year You can enter the day month or year by yping the required number in each field Note the month is entered as a number not a string Clicking on the displays a date picker The date picker displays a month and any day within the month can be selected by clicking on that day The m arrow keys move down or up one month respectively Real fields There are two types of real fields i e either with or without spinners
29. time series button opens a graph window as shown in Figure 23 that allows the user to graph observed flow and model outputs The parameters that are output are selectable from the Edit Recorded Time Series menu option 26 rainfall runoff library Issued 3 Nov 03 Using RRL F mspect Recorded Time series 000000000 L OX Time Series Runoff Obs Runoff Runoff Add Remove M Show Dates Show Graticule iv Show Tips Clear Graph Figure 23 Model output graph window The calibration results button display a list of the optimization results for the latest optimized calibration The data statistics button displays a statistical comparison between observed and simulation runoff for the calibration and validation period An example statistical comparison is shown in Figure 24 BEP Statistics Comparative statistics for runoff Variable Start End Nb Total mm Mean mm Std Dev Skew mm Nash Sutcli R 2 Calibration 11 04 2010 3 02 2020 3586 3343 6036 0 9324048 1 35489747 0 7705695 0 2792912 0 7080838 Verification 19 07 2020 31 12 2025 1992 1627 9641 0 8172510 1 2365736 0 8069591 0 2129435 0 6537946 Data Statistics Variable Start End Length Missing Total mm Mean mm Std Dev Skew mm Runoff 1 01 2010 3 02 2020 3686 0 3554 8238 09644123 1 3736572 0 7040976 Runoff veri 18 02 2020 31 12 2025 2144 0 18125521 0 8454068 1 2244464 0 8003059 Figure 24 Observed and simulated data comparison
30. 002B hasCustomLabel gt false lt Unit_x002B_ hasCustomLabel gt a3 RefUnit a3 BaseUnit id ref 37 xmlns a3 2 http schemas microsoft com clr nsassem TIME Core Units TimeCore 2C 20Version 3D1 0 1390 33501 2C 20Culture 3Dneutral 2C 20PublicKeyToken 23D1 7delc0433b5448b lt dimensions href ref 16 gt refUnit href ref 33 gt scalingFactor 1 scalingFactor offset 0 offset X label id ref 43 m label lt _hasCustomLabel gt false lt _hasCustomLabel gt SimpleUnit x002B dimensions href ref 16 gt SimpleUnit x002B refUnit href ref 33 gt SimpleUnit x002B scalingFactor gt 1 lt SimpleUnit_x002B_ scalingFactor gt SimpleUnit x002B offset gt 0 lt SimpleUnit_x002B_ offset SimpleUnit x002B label href ref 43 gt SimpleUnit x002B hasCustomLabel gt false lt SimpleUnit_x002B_ hasCustomLabel gt Unit x002B dimensions href ref 16 gt Unit x002B label href ref 43 gt Unit x002B hasCustomLabel gt false lt Unit_x002B_ hasCustomLabel gt lt a3 BaseUnit gt lt a3 BaseDimension_x002B_Length id ref 38 xmlns a3 http schemas microsoft com clr nsassem TIME Core Units TimeCore 2C 20Version 3D1 0 1390 33501 2C 20Culture 3Dneutral 2C 20PublicKeyToken 23D1 7delc0433b5448b lt name id ref 44 Length name BaseDimension x002B name href ref 44 gt XElementaryDimension x002B name href ref 44 gt Dimension x002B name href ref 44 gt a3 BaseDimensio
31. 01 1902 12 00 00 AM validationStartDate 31 12 1917 12 00 00 AM validationWarmupToDate 31 12 1917 12 00 00 AM validationEndDate 31 12 1917 12 00 00 AM validationMaxDate 31 12 1917 12 00 00 AM CalibrationType TIME Tools Optimisation ShuffledComplexEvolution DataScaling evapotranspiration panFactor 11111111111 rainfall panFactor 11 11 1 Ad cx dx d COMMENTS Figure 47 Example of rainfall runoff library job file format 8 2 7 QDNR SILO daily time series format txt An QDNR SILO daily time series format file is an ASCII text file that contains daily time series data There is no header line in the file The data is organized in five columns seperated by spaces The first four columns are the date in year month day and julian day format The fifth column is the data value An example of this format is shown in Figure 48 2001 12 29 363 0 2001 12 30 364 O 2001 12 31 365 10 2 2002 1 1 0 2002 1 2 2 0 20021330 2002 14416 Figure 48 Example of QDNR SILO daily time series format 8 2 8 XML times series format tsx The XML time series format is not described An example of the XML format is shown in Figure 49 lt SOAP ENV Envelope xmlns xsi http www w3 org 2001 XMLSchema instance xmlns xsd http www w3 org 2001 XMLSchema xmlns SOAP ENC http schemas xmlsoap org soap encoding xmlns SOAP ENV http schemas xmlsoap org soap envelope xmlns clr http schemas microsoft com soap
32. 01 2010 Max 31 12 2025 Run from 017 January 2010 to 017 January 2010 Run Output Desired runoff units mom Save Runoff as Save Monthly Runoff as View Recorded Time Series In pz E Show Graticule 2 Specily the period of time that you want model results by sliding the start and end date sliders Select the run button From the desired runoff units select what ouput units required To save as daily or monthly data select the appropriate button oO a h CO The Windows save as dialogue will be displayed 65 Rainfall Runoff Library User Guide Save Data As Save in Data 1 E3v SDLWC Mekong My Recent nsw Documents nT 3 ap Desktop My Documents My Computer e My Network Places Save as type Tarsier Daily Time Series tts v Cancel LA Note the default directory for saving data is the RRL data directory The Save as type lets the use save the uutput in various formats 1 Select the output format required 2 Enter a file name without an extension 3 Select the save button 7 2 Saving the project file To save the project file for future use 1 Select the File Save or File Save as menu option 2 The standard windows save dialogue will be displayed Note the default directory for saving project files is the RRL project directory The save as option lets the user save the file under a dif
33. 85 to 31 12 2000 Interval Daily Year 1985 Factor 0 1E 01 01 02 03 04 05 06 30 34 Total Jan 0 0 0 0 165e 97 3070 2509 10956 Feb 2040 1666 1337 1081 1 870 1 Mar 545 450 358 286 217 216 1987 7309 45937 Apr 6910 5528 4520 4818 4947 3982 9081 219146 Dec 33 32558 31655 30729 29827 29154 19364 19051 767745 14627179 Figure 46 Example of IQQM daily time series format 8 2 6 Rainfall runoff library project file jobf The RRL project file is aimed at facilitating the reproducibility of the results obtained during a calibration They are readable ASCII text files with a dual purpose The software uses it to store important settings to keep from one session to another They can also be printed for inclusion in a technical report They should not be edited by a text editor as this is likely to prevent the software from read the file correctly From a technical point of view a text format as opposed to a binary format has been chosen to facilitate the backward compatibility in future versions 70 rainfall runoff library Issued 3 Nov 03 Data storage and file formats A project file as shown in Figure 47 is organised in a series of sections identified by an identify cation string contained between brackets The header information describes the filename and path the creation date and time and user name of the last person creating this project Note comment lines begin with a colon The file has 11 sections 1 Model
34. Rainfall Runoff Library User Manual CONTENTS 1 Iniroduch om ente bss 1 1 1 SUS A bd eM Restu ate tstomemanee T 1 Vil PUFPOS E conri tien decana 1 US SEES A rcc 1 1 2 ara EE EAEE EE ERE ite seaca e t tuat a E Esa ob tae ean 2 Wel 5 A A O pire eec pini 2 US A A NN 2 EL AU tE dada 3 1 3 Data requirements oerte err Oen Ka t e tora EE EXE EXER E ONSA tir 3 Use A eec de eie cerdo aided oed edens 3 13 2 Predicted delta o go 3 1 4 Product componen S usus ces es ptu O RE 3 1 5 References and Training ns eo yep ori er RR SE FED a UR du bed qo yes 3 2 stall atio NR EE OO 5 2 1 Technical specifications cos nn FERE AUR LEE 5 2 151 A e E pasate 6 2 2 DISTIDU ON MEA A a 6 2 241 e A manera edu Meat po 6 2 2 2 Data ein e ED eet hedge it AS 6 2 3 Directori S iii A exe oec de thee lex baad A 6 2 4 SA rect ele Deb evene ad eon REC E RO rte D ORE RAS ORNA 7 2I Sind alone PG o eene repo etx ti ic 7 2 4 2 File server network ie e et eei PUR ea HERREN 8 2 5 User Interface mirin Gta bulles adi eet E dede et etude 8 2 5 1 Windows uias edet oet e D PR bd mero I n ee e RE ORTU 8 2 9 2 Men s itae opto tcm eisdem oO ES irre eadera 9 2 5 3 Toolbar BUTONS sissen as 10 LOA AS UM C nS 10 2 5 5 REDWING WS i ce 12 2 5 6 Mouse amp Keyboard ControlS oooccconnocnccconcnccccconononcnonononacicnnno 15 2 6 Uninstall RR s cestode ROPA tea ee 16 3 USING URK Ei 3s ota dene anrden aa ee ee een noes 17 3 1 Getti
35. Simulation start and end dates output file and flow graph display window Model Input Dates Calibration Sensitivity Simulation Figure 15 RRL Tabbed dialogue Model dialogue The model dialogue as shown in Figure 16 allows the user to select the model type that is to be used for rainfall runoff modelling The user may chose from the following models AWBM Sacramento Simhyd SMAR SWAT core Tank This dialogue has three fields the model select schematic and description When a model is selected the model schematic and description fields will be updated Model Input Dates Calibration Sensitivity Simulation Select Model Sacramento SimHyd SMAR Tank Model Currently using model AV BM p RAINFALL pun 7 rainfall exc surface storages The AWBM is a catchment water balance model that can relate runoffto rainfall with daily or hourly data and calculates losses from rainfall for flood hydrograph modelling The model uses 3 surface stores to simulate partial areas of runoff The water balance of each surface store is calculated independently of the J m 1 A 1 Figure 16 Model selection dialogue 20 rainfall runoff library Issued 3 Nov 03 Using RRL Input dialogue The input data dialogue as shown in Figure 17 allows the user to specify the times series files to be used for rainfall runoff modelling The dialogue has six fields 1 File browser 2 Time seri
36. a generic calibration Note that if the Dynamic Update option was previously not selected checking it will also display the latest recorded graph if any is present Parameter sensitivity It is important to understand how sensitive a model is to certain parameters This is useful to understand how the model functions and also what parameters need more attention than others If the model is significantly affected by a particular parameter than the focus of calibration should be on that parameter The uncertainty of the model will also be closely related to the uncertainty in estimating the most sensitive parameters The RRL provides a facility to investigate the sensitivity of all model parameters Note that in most rainfall runoff models the behaviour of many parameters is related to the values of other parameters i e the models are non linear Consequently the sensitivity of particular parameters may be different depending upon the values of other parameters 63 Rainfall Runoff Library User Guide 7 Saving results This chapter discusses saving results of the model calibration and the project file 7 1 Saving model calibration results To save model results 1 Select the simulation tab dialog as shown below 64 rainfall runoff library Issued 3 Nov 03 Saving results Model Input Dates Calibration Sensitivity Simulation Tools Selecta standard plotting option Choose astan Clear Plot Input Data Dates Min 1
37. ach layer holds 25 mm at field capacity 4 Rainfall Runoff Library User Guide The dimensionless parameter H is used to estimate the variable H the proportion of rainfall excess contributing to the generated run off as saturation excess run off or the Dunne run off H is obtained as a product of H rainfall excess and soil saturation Soil saturation is defined as the ratio of available soil moisture in mm at time t days and 125 mm representing the maximum soil moisture content of the first five layers The parameter Y mm d represents the infiltration capacity of the soil and is used for estimating the infiltration excess run off Hortonian run off The dimensionless parameter T is used to calculate the potential evaporation from pan evaporation E Generated surface run off is calculated from the excess rainfall rainfall minus potential evaporation as saturation excess run off shallow sub surtace flow plus the Hortonian run off and plus a proportion 1 G of moisture in excess of the effective soil moisture storage capacity g i e through flow The remaining proportion G of the latter i e the deep drainage component discharged from the groundwater system to the stream is routed through a linear reservoir and the total generated surface run off is routed using a gamma function model form to obtain the daily total estimated discharge at the catchment outlet 4 4 2 Routing Groundwater and surface run off
38. an rainfall runoff library Prepared by Geoff Podger CRC for Catchment Hydrology Version 1 0 November 2003 www toolkit net au rrl ar MODELLINGTOOLKIT 2 supported by the CRC for Catchment Hydrology CATCHMENT Document History Date Author Revision Description of Change 03 11 2003 Geoff Podger 1 0 0b First version of the User Guide Copyright notice CRC for Catchment Hydrology Australia 2003 Legal Information To the extend permitted by law the CRC for Catchment Hydrology including its employees and consultants accepts no responsibility and excludes all liability whatsoever in respect of any person s use or reliance on this publication or any part of it TOC Acknowledgements This User Guide has been written and edited by Geoff Podger The conversion of the User Guide to on line help was done by Nick Murray The User guide was reviewed by Tanya Jacobson The software was developed by Jean Michel Perraud Rainfall runoff models included in the library were provided by Walter Boughton Francis Chiew and Geoff Podger The software graphic images were prepared by Susan Daly The Product Manager for the Rainfall Runoff Library is Geoff Podger Department of Infrastructure Planning and Natural Resources CRC for Catchment Hydrology 2 CSIRO Land and water CRC for Catchment Hydrology 3 Melbourne University CRC for Catchment Hydrology 4 CRC for Catchment Hydrology TOC
39. arge from the groundwater system e Discharge to the stream until a maximum threshold after which discharge to land occurs following shallow watertable development Discharge to the land surface that is locked in the landscape and is eventually lost to the atmosphere Inter basin transport from the local groundwater system to the regional groundwater system Two assumptions are made in treating the groundwater routing components as a single linear reservoir Discharge to the land that does not eventually reach the river is negligible Inter basin transport from the local flow system to a regional groundwater system is substantially less than the discharge to the stream Bear 1979 Therefore QF T is comprised mainly of the groundwater discharge to the stream and to the land surface that eventually reaches the stream The lag times between natural replenishment and groundwater discharge are substantial and the groundwater system behaves like a highly damped system This mechanism can be visualised as one of displacement whereby water from episodic drainage events is continually added at the bottom of the root zone and is removed from the groundwater system at a very slow rate This process can be expressed by a single linear reservoir with a large storage coefficient K 43 Rainfall Runoff Library User Guide 4 5 4 6 The pulse response function for the groundwater component can be obtained in a manner analogo
40. ation of which is detined by the user during installation Help contains the html and hlp files for RRL Help Documentation contains two subdirectories User Guide and Training The User Guide subdirectory contains this document and the Training subdirectory contains training documentation rainfall runoff library Issued 3 Nov 03 2 4 Using RRL Projects contains 6 project files one for each rainfall runoff model in the RRL Data contains project data This contains 6 subdirectories one associated with each of the 6 projects in the project directory The subdirectory structure of the installed RRL software is discussed in section 8 1 __ Program Files Adobe Common Files EMSS E Assistant Bin E Data Auxillary Calibration ClimateSets DiffuseTreatments Flow Images LandUseMaps LinearTreatments PointSource E Projects E 2 r_99_00 Current Future PreSettlement Renames StorageData Installation 2 4 1 Stand alone PC Atter downloading the software to a local directory double clicking setup exe will initiate an install wizard which leads you through the install procedure Once again you must indicate that you have read understood and agree to be bound by the RRL licence agreement There is only one decision that needs to be made during the installation and that is the name of the install directory default is C Program Files toolkit Ra
41. braic development of linear modelling techniques Journal of Hydrology 133 17 40 Lewis M L and Trosset M W 2000 Direct search Methods Then and Now NASA ICASE Report No 2000 26 Lewis M L Torczon V and Trosset M W 2000 Direct search Methods Then and Now NASA Langley Research Center ICASE Report No 2000 26 Nash J E 1960 A unit hydrograph study with particular reference to British catchments Proceedings of the Institute of Civil Engineers 17 249 282 Nash J E and Sutcliffe J 1970 River flow forecasting through conceptual models Part 1 A discussion of principles Journal of Hydrology 10 282 290 Nash J E and Barsi B I 1983 A hybrid model for flow forecasting on large catchments Journal of Hydrology 65 125 137 O Connell P E Nash J E and Farrel J P 1970 Riverflow forecasting through conceptual models part 2 the Brosna Catchment at Ferbane Journal of Hydrology 10 317 329 Rosenbrock H H 1960 An automatic method for finding the greatest or least value of a function Computer Journal 3 175 184 Singh V P Computer Models of Watershed Hydrology Water Resources Publications Colorado USA 6 165 214 80 rainfall runoff library Issued 3 Nov 03 References Tan B Q and O Connor K M 1996 Application of an empirical infiltration equation in the SMAR conceptual model Journal of Hydrology 185 225 295 Tuteja N K and Cunnane C 1999 A quasi phys
42. crosoft com clr ns System Collections gt lt LoadFactor gt 0 72 lt LoadFactor gt lt Version gt 0 lt Version gt lt Comparer xsi null 1 gt lt HashCodeProvider xsi null 1 gt lt HashSize gt 11 lt HashSize gt lt Keys href ref 17 gt lt Values href ref 18 gt lt a2 Hashtable gt lt al EvenTimeStep id ref 9 xmlns al http schemas microsoft com clr nsassem TIME DataTypes DataTypes 2C 20Version 3D1 0 1392 2858822C 20Culture 3Dneutral 2C 20PublicKeyToken 3D5 9b30b0da8a82a84 gt lt _span gt POYOM1DTOHOMOS lt _span gt 74 rainfall runoff library Issued 3 Nov 03 Data storage and file formats lt al EvenTimeStep gt lt SOAP ENC Array id ref 10 SOAP ENC arrayType xsd double 6209 gt lt item gt 0 lt item gt lt item gt 0 lt item gt lt item gt 0 lt item gt lt item gt 10 2 lt item gt lt item gt 0 lt item gt lt item gt 0 lt item gt lt item gt 16 lt item gt lt SOAP ENC Array gt lt SOAP ENC Array id ref 11 SOAP ENC arrayType xsd anyType 0 lt SOAP ENC Array gt lt SOAP ENC Array id ref 12 SOAP ENC arrayType xsd anyType 0 lt SOAP ENC Array gt lt a2 ArrayList id ref 13 xmlns a2 http schemas microsoft com clr ns System Collections gt lt _items href ref 19 gt size 2 size version 2 version lt a2 ArrayList gt lt a3 UnitPrefix id ref 14 xmlns a3 http schemas microsoft com c
43. d by pre specified starting points This method works by dividing the parameter values into a specified number of increments between the specified bounds For each of these possible starting points a pattern search is carried out The best optimum of the pattern searches is taken as the global optimum Rosenbrock method The Rosenbrock method Rosenbrock 1960 Lewis 2000 is a local search method bearing some similarities with the Pattern Search method described in the preceding section This search method returns at each step a point at least as good as the previous one in the parameter space lt was originally designed in order to optimise an industrial process in chemical engineering and to handle response curves with peculiar features such as functions with narrow curved valleys e g Rosenbrock s banana function detailed in this section The two main improvements over the pattern search are a better use of the local information from the response curve surrounding the point in the parameter space and an adaptive step size This search method proceeds in a series of so called stages A stage is the search in the parameter space following successive directions along an orthonormalised set of vectors base of the same dimension as the parameter space To illustrate the idea behind the algorithm we will step through the first few iterations of a Rosenbrock search on the aforementioned banana function rather than a set of equations This two
44. d check box fixing that can optional fix the parameter in optimised calibration Note the parameters that are displayed in this dialogue are dependent on the rainfall runoff model selected 24 rainfall runoff library Issued 3 Nov 03 Using RRL E Manual Calibration Model details os 1 Fixed Mex 1 Fixed KSur 0 35 Min KBase 0 95 Min Mex Infinity Fixed Max Infinity Fixed Max 1 Fixed NI Figure 21 Calibration parameters C2 Min Min og g ca 19 min 0 Mex Infinity Fixed n0 3 Min There are three tabs on the calibration tab dialogue The three tabs are 1 Generic 2 Custom and 3 Manual The generic tab contains drop down menus and buttons to access options for the automatic calibration of model parameters The user can select optimisation method primary and secondary objective functions and data transformation By clicking the mouse on the optimisation method drop down list the user can select from the following optimisation methods e Uniform random sampling e Pattern search e Multi start pattern search e Rosenbrock search e Rosenbrock multi start search e Genetic algorithm e Shuffled Complex Evolution SCE UA The optimisation parameters can be set by clicking on the parameter Note the parameters that are displayed are dependent on the optimisation method that is specified By clicking the mouse on the primary objective function drop down list the user
45. d t e DEENA 10 Faure S Popup lists miro ob eb c testen us ee n e ERR Per D EUER IN IRE Uo co dogs 10 Figure 4 Graph window ucc rite eee a o ette IR Pese e Eta Pe eeu e baeo uns 12 Figure 5 GraphipopUpss siet ER CREE nece t e Be te E ud Uus e due Mus e da 12 Figure 6 Graph properties dialogue oa Leere uere e e eerie tee Eta 14 Figure 7 Colour picker de de dee Re Pe dE RET Ra UHR d RR 14 Figure 8 Line style selections ies e te e ote ehe eie ee e ene ed 15 Figure 9 RRL Atop level a da 17 Figure TO File drop down menu d deer ete tere tee e re etae e eee gi 18 Figure 1 Editdrop down menses ett i e et a ree PEERS OR REF nadie a 18 Figure T2 View drop down merui eise a nani es 19 Figure 13 Tools drop down menu ic este ate coe ae t b Re rep he e rupe e de 19 Figure 14 Helpdrop downmenu A e eder rte e p oe td e t re Eee 19 Figure ERRE Tabbed dialogus skanna t REESE e ee te PR tele es 20 Figure 1 6 Model selection dialogue ssssseee eee 20 Figure 17 Inp tidialogue ceo erc ET Etre ERE EE ECRIRE ERG 21 Figure T8 Dates dialogUe eene bd ee et ret tere eere ere ben eps e re Re pe xa ERA 22 Figure 19 Warm up period estimation error dialogue sssss e 23 Fig re 20 Galibration dialogue en e tette nte conten HR PUER EN e to sands 24 Figure 2 T Galibration parameters co tese tte ente ttes tette LY e tet Ponit eese das 25 Figure 22 C stomtabza eet ER ERR is 26 Figure 23 Model output graph window
46. del wil perform well during dry periods but is likely to underestimate flows during wet periods When the model has been calibrated it is worth checking how robust the model calibration is by assessing if soil stores reach the extremes of fully saturated or fully dry This will give some indication of how the model operates under fully saturated surface flow and pure base flow If the runoff match during fully saturated conditions is good then the model will perform 57 Rainfall Runoff Library User Guide reasonably well during wet periods If the baseflow recessions match reasonably well during an extended dry period then the model is likely to perform well in dry periods Quite often during optimised calibrations it is possible for storages to be calibrated to operate incorrectly with too much baseflow coming out fo interflow stores and vice versa It is worth checking that the relative proportions of flow coming out of these stores and the recessions are appropriate Check that the intertlow store empties quicker than the base store Check that the volumes of interflow and base flow are appropriate for the type of catchment 6 2 2 Calibration and validation Given a suitably long period of flow record this is the preferred method for calibrating rainfall runoff models This method gives a way of assessing the robustness of the model for periods outside of the calibration period When using calibration and validation periods it is importa
47. dent upon the type of variable that is being displayed Currently there are only two types of variables supported real numbers and dates For real numbers the precision can also be specified The font used for the label can also be specified with the standard Microsoft Windows font selection dialogue The graph graticules may be turned on and off for either the x or y axes This is done by clicking the mouse on either true or false in the drop down list of the appropriate parameter The line style can also be selected via a drop down menu The available options are shown in Figure 8 DashDotDot Custom Figure 8 Line style selections 2 5 6 Mouse amp Keyboard Controls This User Guide assumes that you have a standard mouse and keyboard configuration The left mouse button is commonly used under Windows to access the normal select or normal drag function The right mouse button is normally used to access context sensitive menus and drag functions These buttons have the same usage in the RRL You may prefer to use keyboard hotkeys or a glide pad 15 Rainfall Runoff Library User Guide 2 6 Uninstalling RRL You uninstall RRL via Settings Control Panel Add Remove Programs Removing the RRL software is only a matter of seconds To ensure complete removal of the software use Windows explorer to browse to the RRL install directory and then remove this directory Caution all existing data and project files that may have been sa
48. dow so it covers the full screen E e After maximising a window click E restore the window to its previous size Note 1 You can also double click the window s title bar to maximize it or restore it to its previous size 2 You can click on the Toolkit icon on the left of the title bar or right click on the title bar to display a menu to restore move size minimise maximise or close the window The title line contains the name of the software followed by a version number The version number is in three parts x y z where x Major Version indicator integer starting at 1 y Minor Version indicator integer starting at 0 may go to any number but rest on a major version change rainfall runoff library Issued 3 Nov 03 Using RRL z Bug fix indicator integer starting at O may go to any number but reset to zero on a minor or major version change The suffix b may be added to a version number to denote that the software is a beta release Major version increment This indicator is incremented following a substantial change to the way the software operates and is usually associated with a significant change to documentation and training material An example of a major version change would be the port of EMSS from Tarsier to TIME Minor Version increment This indicator is incremented when a new feature is included in the model An example is adding a new rainfall runoff model or optimisation method to this library It i
49. e and K is the base flow recession constant of the time step being used daily or hourly The surface runoff can be routed through a store if required to simulate the delay of surface runoff reaching the outlet of a medium to large catchment The surface store acts in the same way as the base flow store and is depleted at the rate of 1 0 KS SS where SS is the current moisture in the surface runoff store and KS is the surface runoff recession constant of the time step being used 36 rainfall runoff library Issued 3 Nov 03 Description of models SURFACE RUNOFF 1 BFI EXCES BASEFLOW A1 A2 A3 RECHARGE ROUTED RE fe BFI EXCES SURFACE RUNOFF 1 KS S BS BASEFLOW 1 K BS PARAMETERS AND STATE VARIABLES 40 y as TOTAL 1 C3 S ge C s CI C urface Storage Capacities RUNOFF Al A3 Partial Areas Represented by Surface Storages BFI Baseflow Index K Daily Baseflow Recession Constant BS Current Volume in Baseflow Store KS Daily Surface Flow Recession Constant SS Current Volume in Surface Routing Store Figure 32 Structure of the AWBM rainfall runoff model 4 2 Sacramento The SAC SMA model represents the moisture distribution in a physically realistic manner within hypothetical zones of a soil column The model attempts to maintain percolation characteristics to simulate stream flow contributions from a basin The components of the SAC SMA are tension water free water surface flo
50. e colours for wettest and driest periods can be changed by clicking the mouse on the colour bar following the buttons This will display a colour palette selection window Note the calibration and verification period colours can be modified in a similar manner The calibration start end and warm up date fields and associated date sliders allow the user to manually specify the respective calibration dates Note the end date must be after the start date and the warm up period must be between the two dates Alternatively the warm up period may be approximated by clicking the mouse on the find appropriate calibration warm up The model determines this by identifying a period where the contents of the storages reach a storage depth that is within 1 irrespective of the value of the model parameters Note under certain conditions the model cannot determine this period in which 22 rainfall runoff library Issued 3 Nov 03 Using RRL case a warning message as shown in Figure 19 is displayed and this period needs to be specified manually Could not find an appropriate calibration warmup please adjust manually Figure 19 Warm up period estimation error dialogue The perform verification date fields remain greyed until the perform verification check box is checked The operation of these fields as well as find appropriate verification warm up work in a similar manner to the calibration fields Note there are no constraints for overlapping of ca
51. e href ref 46 gt a3 BaseDimension x002B Time lt SOAP ENV Body lt SOAP ENV Envelope gt Figure 49 Example of XML times series format 8 2 9 SWAT daily rainfall time series format pcp An SWAT daily rainfall time series format file is an ASCII text file that contains daily time series rainfall data The file has a four line header followed by daily data values The four lines of information contained within the header are 1 The swat file description header 2 Lati followed by the latitude of the site in degrees 3 Long followed by the longitude of the site in degrees 4 Elev followed by the elevation of the site in metres The data is organized in three columns The first column is the year 4 characters followed by the julian day 3 characters The third column is the data value to one decimal place 5 characters An example of this format is shown in Figure 50 Precipitation Input File pcp pcp 20030225 AVSWAT2000 SWAT interface MDL Lati 14 77 Long 102 7 Elev 167 1985001000 0 1985002000 1985003000 1985004000 1985005000 oo0o0oo Figure 50 Example of SWAT daily rainfall time series format 8 2 10 Tarsier daily time series format tts An Tarsier daily time series format file is an ASCII text file that contains daily time series data The file has a 21 line header followed by daily data values The four lines of information contained within the header are 1 The Tars
52. ed by the period of input data specified in the input dialogue There are also two date fields and date sliders to specify the period that the model is to be run over Note the start and end dates will be constrained to the earliest and latest dates Clicking the mouse on the run button will run the rainfall runoff model with the current set of parameter values for the period specified in the output period fields If a graph type is selected in the graph area a graph of the model flow result will be displayed The desired runoff units allow the user to choose the output units that the data are to be saved as There currently three options available mm d m3 s and ML d The output buttons allow the user to save the results of a model run to either a monthly or daily output file Clicking on an output button displays a file save as window as shown in Figure 27 This window allows the user to save the model results in one the TIME supported output formats as described in Chapter 8 29 Rainfall Runoff Library User Guide Save Data As Save in Data ect Ey DLWC Mekong My Recent nsw Documents BNT 3 Sao Desktop My Documents My Computer My Network Places Save as type Tarsier Daily Time Series tts Y Cancel Figure 27 File save as window The view recorded time series button opens a graph window as shown in Figure 23that allows the user to graph observed flow and model outputs The para
53. ent Area 1234 km Comments Thie is an example site Value for missing Data 9999 OK Cancel Figure 31 Catchment information dialogue There are 4 fields that may be entered in this dialogue 1 Location 2 Catchment area 3 Comments and 4 Missing data value The location is a text string that describes the location that is to be modelled by the RRL This field may be left blank The catchment area is a real field that specifies the size of the catchment in km A value must be entered in this field before the ok button is selected This value is used to convert between depth of runoff and flow The missing data value is used to identify missing data in the input data 34 rainfall runoff library Issued 3 Nov 03 3 3 Using RRL Note these parameters can be adjusted within the rainfall runoff menus via Edit Project details When all of the information has been entered the user can click the mouse on the OK field Loading an existing project To load an existing project select File Open Project and a standard windows browse window will appear By default this file will open up in the RRL project directory Once a job file is selected the RRL tab dialogues will be loaded with the saved parameters for the project The time series data will only be loaded if the data exists in the specified directories in the correct format 35 Rainfall Runoff Library User Guide A 4 1 Description of models
54. epara us esie o Sere POL A aa 54 6 1 1 Catchment characteristics sess 54 6 1 2 Rainfall datas icy er an ba Eee omi Meo DR un e E 55 6 1 3 Evapotranspiration data nee ttr t RUN VERE ha ie 56 0 154 Flow datas doeet DNE bn a IO e weet ee ee ena TU 56 6 2 Calibration and validation period esce 57 6 2 1 Calibrating over the entire period ccoooocncccnnoccncconononncnoncnnno 57 6 2 2 Calibration and validation ooooccccccnnocnccnonoccncnononinacinnonanos 58 6 2 3 Model warm Ups e AS 59 6 24 Setting the periods uta ets 59 6 3 Manual calibration of models eese 59 63 4 Dynamite Ubddlis ose ER ba 59 6 3 2 List of calibration parameters occcooooccccccnnncoccnnnononcnononanacinncnnnos 59 6 3 3 Update graphics Bur exis Et 60 6 4 Automatic calibration of models esses eene 60 6 4 1 Custom calibration of AWBM sess 60 0 4 2 Generic calibration A EI EH IM 60 6 5 Parameter sensHvily s ec Eoo pec Cao e e ora OR E A EROR RUD RR CH RUE 63 7 SAVIO FesUlls cito eoa OS 64 7 1 Saving model calibration results eese 64 7 2 Saving the project file cce Gc E AA 66 8 Data storage and file formats 67 8 1 Data STORAGE x oie ad oO ex ay oka idas 67 8 1 1 Installation directory cernere ne dando rene Een eer nee 67 8 1 2 Projects sub directory ous
55. eraged as in equations 2 and 3 to represent the daily values 42 rainfall runoff library Issued 3 Nov 03 Description of models 0 l frar t T 0 1 ere 6 The linear model described by equation 4 is the simplest representation of a causal time invariant relationship between an input function of time generated run off and the corresponding output function routed run off It is used in conceptual modelling as a component representing the routing or diffusion effects of the catchment on those components of the rainfall hyetograph contributing to the outflow Qj 3 Hr jy T where m memory of the pulse response function d The parameter pair n and nK are chosen for optimisation rather than n and K separately because n is a shape parameter and nK is the scale parameter Expressed in this way the two parameters are likely to be more independent than would be n and K separately both of which contribute to the scale and to the shape although in different ways Groundwater routing component The mass balance equation for the groundwater system can be written as in equation 5 dS r dt r r Qf r DS r 5 1 where or recharge to the groundwater system MM S the groundwater system MM see T time s S T storage of the groundwater Of discharge from system mm and D d dt is the differential operator s J There are three basic components of disch
56. es rainfall 3 Time series evapotranspiration ET 4 Time series flow 5 Input statistics button 6 Reset all inputs button Model Input Dates Calibration Sensitivity Simulation E C Program Files Toolkit Rainfall Runo Drag the appropriate data form the tree view onto the pertaining model input AWBM sample data DL WC NSW NT QLD junk iqqm Calibration Input HE m Rx ronda 91 4188701 1 902011 918011 930111 9540111 9670111983 105 11958 01 1963 01 19711 01 1973 01 1987 gt Input statistics Reset All Inputs Figure 17 Input dialogue The file browser is used to locate a file that may then be dragged and dropped onto the appropriate time series field The file that is selected must be in a valid times series format as discussed in Chapter 8 The units of the data must be compatible with the field that the data is dropped on The rainfall and evaporation fields are expecting height units while the flow field is expecting flow or height units The units of the flow input data can be specified by the drop down menu in the flow input field Note that both rainfall and evapotranspiration data must be entered as millimetres When an appropriate time series data file is dropped on the time series field then the data will be displayed in that field The input statistics button displays summary statistics for each set of time series dataWhen the reset all inputs button is selected all of the time series data field
57. f 3 Jt Rainfall Runoff Library Beta Projects TANK_JARDINE jobf 4 JTookit Rainfall Runoff Library Beta Projects simHyd jobf Exit Figure 10 File drop down menu An RRL project file is a text file that contains information about the configuration of a rainfall runoff model including e Model type e Location e Input data e Parameter e Calibration dates and methods e Dota scaling factors Note 1 The close project save project save project as and print project options will be greyed until a project file is created or opened 2 The previously opened file option will only display if previously opened files were specified and still exist If no files exist then no menu options will be displayed Edit drop down menu The Edit drop down menu is shown in Figure 11 This menu allows users to edit the RRL project file and set default preferences ET View Tools Help Figure 11 Edit drop down menu Note that all options will be greyed if a project file has not been opened or created 18 rainfall runoff library Issued 3 Nov 03 Using RRL View drop down menu The View drop down menu is shown in Figure 12 This menu allows users to view the RRL project file and default preferences view Tools Help Figure 12 View drop down menu Note that all options will be greyed if a project file has not been opened or created Tools drop down menu The Tools drop down menu is shown in Figure 13 This menu allo
58. fall sites Difference in average annual rainfall as compared to the catchment average annual rainfall Proximity to the catchment Correlation with flow peaks The number of sites used If the difference in average annual rainfall is great e g more than 20 then the rainfall process for the catchment and selected site are probably quite different and this is not a good station to use The RRL provides a data scaling dialogue as shown in Figure 41 that allows monthly or annuall factors to be applied to rainfall data to adjust for catchment average rainfall choose multiplier to apply to Model Inputs IL Ox Jan Feb Mar Apr May Jun Jul Aug Sept Oct Nov Dec r foo foo fo fho fi 00 oo foo Afio foo Afio Hio io A r ho froo Jfroo Jf oo io roo foo o0 fi 00 foo Hho Hio 4 Figure 41 Rainfall and evaporation data scaling Unfortunately studies have shown that rainfall decorrelation distance is approximately 10km and there are not many places in Australia where rainfall stations are this close together In most cases stations in the catchment should have priority over ones outside the catchment However in quite a lot of cases there may not be any long term stations in the catchment In 95 Rainfall Runoff Library User Guide cases like this short term stations in the catchment may be used to assess which long term stations are most representative A good method for assessing how well a
59. ferent name 66 rainfall runoff library Issued 3 Nov 03 8 1 Data storage and file formats Data storage and file formats Data storage The directory structure of the RRL is organized relative to the installation directory root The default root directory is c program files toolkit rainfall runoff library The data storage structure is Installation directory Projects Data Help Example AWBM Example Sacramento Example Simhyd Example SWAT Example SMAR Example Tank Rainfall Evaporation Flow Documentation 8 1 1 User Manual Training Installation directory The installation directory contains the executable code and associated dll s required to run the RRL 67 Rainfall Runoff Library User Guide 8 2 8 1 2 Projects sub directory The project sub directory is the default directory used to store the RRL project files Initially this directory contains a project file for each of the rainfall runoff models 8 1 3 Data sub directory The data sub directory contains an example directory for each of the rainfall runoff models These example directories contain the input data files necessary to run and calibrate the models They are also associated with the example project files contained in the project directory The data directory also includes three data directories for rainfall evaporation and flow data collected from various sources 8 1 4 Help The help directory contain
60. ff 1 3 2 Predicted data The model outputs daily and monthly flow or depth of runoff Units Time scale Climate Rainfall mm d Daily Evapotranspiration mm d Daily Flow Observed flow mm d Daily ML d m3 s Simulated flow mm d Daily or ML d monthly m3 s Table 1Data available for interrogation Product components The RRL package includes e ihe RRL software and sample data e User Guide and e Sets of workshop exercises References and training The RRL is a product developed by the CRC for Catchment Hydrology CRCCH Rainfall Runoff Library User Guide As part of the product delivery the CRCCH runs training workshops Details of workshops are posted on the Toolkit web site www toolkit net au in the news and events sections References for relevant conference papers and journal articles are also available at the Toolkit web site in the RRL member s area rainfall runoff library Issued 3 Nov 03 2 2 1 Using RRL Installation This Chapter of the User Guide covers hardware and software requirements to run the RRL the contents of the distribution disk or download file including the licence agreement the directories created by the installation the installation procedure for the RRL and its data how you access RRL once it is installed on your computer how to uninstall RRL from your computer Technical specifications Minimum Requirements Processor 133 megahertz MHz Intel Pe
61. generated from the water balance component are routed to simulate the associated lags between rainfall events and flow out of the catchment The governing equations used in routing component of the SMAR model are presented as follows Kachroo and Liang 1992 The surface runoff routing component The surtace run off generated from the landscape is routed attenuation and lag to the catchment outlet using the linear cascade model of Nash 1960 The model was obtained as a general solution relating a given input of unit volume to a given output as in equation 1 le IN YU OS 2 Tu ol K IB de where f simulation time step d T time s K 2 K K K are the storage coefficients of n linear reservoirs in cascade A t ordinates of the pulse response function d and and n exp 0 c ac is the in complete Gamma 0 function dimensionless It was shown by Nash 1960 that under constraints of conservation stability high damping and the absence of feedback this two parameter equation with n an integer and K positive is almost as general a model as the differential equation of unlimited order With additional flexibility obtained by allowing n to take fractional values the impulse response of this equation has the ability to represent adequately almost all shapes commonly encountered in the hydrological context The generated run off r mm d and the routed run off Oj mm d can be time av
62. h fields except the Show Graticule check box and Clear Plot button will be greyed until the model is run Once the model is run the Select a standard plotting option field will be activated and the required graph type can be displayed by selecting from the list of graph types in the drop down list The graph plotting options are e Daily time series e Monthly time series e Daily flow duration The graph axis options are dependent upon the graph type that is selected and are listed in Table 6 33 Rainfall Runoff Library User Guide 3 2 Graph type X date X log X probability Y log Daily time series Yes No No Yes Monthly time series Yes No No Yes Flow duration Yes No Yes Yes Table 6 Calibration graph axis options The x date when checked displays dates on the x axis and when not checked it is days since start date The x log when checked plots a log x axis and when not checked it is a linear x axis The x probability when checked plots a probability x axis when not checked it is a 0 100 sale x axis The y log when checked plots a log y axis and when not checked it is a linear y axis The current graph can be cleared by clicking the mouse on the clear plot button Creating a new project To create a new project select File New Project and a Create new project a catchment information dialogue will be displayed as shown in Figure 31 E Catchment Information Location Winding River at Middle gauge Catchm
63. hen running the model and assessing the objective function This repeat for a specified number of times and the option with the best objective function value is taken as the optimum solution Pattern search The pattern search is the simplest of all the search methods and has the advantage that it is quick but can suffer from finding local optimums rather than global optimums This is particularly the case when models are strongly non linear The problems of reaching local optimums can be overcome by using a multi start on the search as discussed in the following section The search works by the following method Start with an initial value and search increment for each of the parameters 2 Evaluate the objective function for an incremental increase and decrease in current value 3 If the objective function improves in one direction set the parameter to that value 4 Increment each of the parameters in the optimum direction and evaluate the objective 5 Repeat to step 4 until there is no improvement in any of the parameters 6 Halve the incremental and go to step 2 until the number of specified interval halvings is reached Note if at any stage a parameter reaches a boundary the parameter is limited to the specified boundary value 47 Rainfall Runoff Library User Guide 5 3 5 4 Multi start pattern search The initial sampling of the parameter space provides the potential for locating the global optimum without being biase
64. ian water Groundwater which rises above the ground surface under its own pressure by way of a spring or when accessed by a bore Bedrock A general term for the rock usually solid that underlies soil or other unconsolidated material Bore well Drilled or dug below the surface to access water in an aquifer system Discharge The process by which water leaves the aquifer Drawdown A lowering of the water table of an unconfined aquifer or a reduction of the water pressure in a confined aquifer Drawdown is the result of pumping of groundwater from wells Fractured rock aquifer These occur in hard rocks that have been subjected to disturbance deformation or weathering creating joints bedding planes and faults which water can flow through These are a major source of dryland salinity in the tablelands and slopes of NSW Groundwater The water contained in interconnected pores in rock and alluvium below the water table in unconfined aquifers and anywhere in confined aquifers Groundwater system Another term for an aquifer ydraulic conductivity The rate at which water flows from one point to another through an aquifer ydrogeological Those factors that deal with subsurface waters and related geological aspects of surface waters mpermeable layer A layer of material which does not allow water to pass through Observation well bore A non pumping well used to obse
65. ical snowmelt run off modelling system for small catchments Hydrological Processes 13 12 13 1961 1975 Tuteja N K Beale G T H Summerelle G and Johnston W H 2002 Development and validation of the Catchment Scale Salt Balance Model CATSALT Version 1 Department of Land and Water Conservation Centre For Natural Resources Report ISBN O 7347 5244 X Wang Q J 1991 The Genetic Algorithm and its Application to Calibrating Conceptual Rainfall Runoff Models Water Resources Research Vol 27 No 9 2467 2471 81 Rainfall Runoff Library User Guide 10 Glossary 10 1 Surface water Hydrology Acronym Term Definition Baseflow The component of streamflow that originates from groundwater and supports streamflows during long periods of no rainfall Catchment The area of land drained by a water course that could range from a small runnel to a creek or a river In hydrologic terms every point in the landscape is the outlet of a catchment of some description As the term has a wide range of applicability it gets used very broadly and loosely The equivalent term in the US is watershed Critical drought The duration in time that resulted in the either the lowest dam levels aquifer levels or pressures or longest duration of restricted water access Deep drainage Water in the soil that percolates down below the root zone of plants ie a component of infiltration water and therefore cannot be t
66. ier version number header Reference to author of Tarsier File path and name 2 3 4 Name of software used to crete the file 5 Date and time file was created 6 Tarsier timer series data class TTimeSeriesData 78 rainfall runoff library Issued 3 Nov 03 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Data storage and file formats File version number Number of header lines set to 1 l Number of daily data entries in the file Xlabel is always Date Time for time series data Y Label Y1 fixed field Y2Label Y2 fixed field Data units Grid position east in metres Grid position north in metres Latitude followed by the latitude of the site in degrees Longitude followed by the longitude of the site in degrees Elevation followed by the elevation of the site in metres Header character The data is organized in four columns separated by spaces The first column is the year followed by the julian day The third column is the data value The fourth column is a data quality code Is ok and 4 is missing An example of this format is shown in Figure 51 Tarsier modelling framework Version 2 0 Created by Fred Watson File Name C tmp TSFormat precip tts Generated from TIME Framework Date 24 10 2003 3 21 30 PM File class TTimeSeriesData FileVersion unknown HeaderLines 1 T NominalNumEntries 6209 XLabel Date Time YlLabel Y1 Y2Label Y2
67. infall Runoff Library Potential Problems If your computer doesn t have enough space to install the data the install procedure will give you an out of disk space message Select Cancel to exit the installation procedure You will need to find a larger disk drive If the net framework is not installed on your machine the install procedure will give you a warning message Select cancel the installation procedure You will need to install the NET redistributable software as described in Section 2 11 Rainfall Runoff Library User Guide 2 5 2 4 2 File server network The RRL is not designed to be used on a file server network It is intended that it be used on a stand alone PC User Interface 2 5 1 Windows TIME uses standard Microsoft Windows windows as shown in Figure 1 The features in the window are the e Title bar that includes the Toolkit icon followed by the product name and a version number e The standard windows minimise maximise restore and close icons e Optional top level menu Rainfall Runoff Library v1 0 0 File Edit View Run Tools Window Help MODELUNGTOOLKIT Figure 1 Example of standard window To minimise or maximise a window or restore it to its previous size Click the appropriate button in the upper right corner of the window e Click m minimise the window to a taskbar button To restore the minimised window to ifs previous size click its taskbar button e Click m maximise the win
68. ing probability value for the worst point p1 is then 2 C m To ensure that all probability values are nonnegative C lt 2 C 2 was used 6 Select two points A and B from the m points at random according to the probability distribution pj 1 2 m 7 Select two bit positions k1 and k2 along the overall coding of the parameter set at random giving each bit position the same chance If kl gt k2 their vales are exchanged 8 Form a new point by taking the values of the bits trom k1 to k2 1 of the A point coding and the values of the bits k2 to the and from 1 to k1 1 of the B point coding 9 Occassionally change some of the bits of the newly formed point A bit value O will become 1 and vice versa This occurs to each bit only ata a very small probability pm pm 0 01 was used 10 Repeat steps 5 8 m times so that m new points are produced The original m points are then replaced by the new ones forming a new data base for further search 1 Repeat steps 2 9 The best point found so far is always recorded Termination of the search is specitied by a total number of function evaluations Steps 6 and 7 form the core of the method Better points have a better chance to be chosen to form new points This is analogous to the survival of the fittest in the theory of natural selection The better performing individuals produce more offspring A new point is formed by taking different blocks of bits from the codings of the two original po
69. ints This is analogous to crossover in the theory of genetics An offspring takes some of the genes from one parent and some from the other one Fit parents are likely to produce fit offspring The combination of selection and reproduction improves the performance level of the population as the process move on The occasional change of bit values in step 8 is analogous to mutation in the theory of genetics It provides some background variation Shuffled complex evolution The shuttled complex evolution SCE UA method is based on a synthesis of four concepts Combination of deterministic and probabilistic approaches 2 Systematic evolution of a complex of points spanning the parameter space in the direction of global improvement 52 rainfall runoff library Issued 3 Nov 03 3 4 Description of optimisers Competitive evolution and Complex shuffling A general description of the steps of the SCE UA method is given below a more detailed presentation of the theory underlying the SCE UA algorrithm has been given by Duan st al 1992 1993 1 Generate s sample points randomly in the feasible parameter space and compute the objective function value at each point Rank the s points in increasing objective function value such that the first point represents the smallest objective function value assuming that the goal is to minimise the objective function Partion the s points into p complexs each containing m point
70. ion of the standard one dn o y LoD mean obs y where A is between O and 2 This simple box cox transform allows giving more weight to the lower flows if less than 1 or to the high flows more than 1 in the calibration process Sum of the squares of error gt m i o Root mean squared error JN m oi n where n is the length of the time series 61 Rainfall Runoff Library User Guide RMS about the bias Y ma o n The biais is defined as bias and the root mean squared about the bias as Y Gn oG bias n Sum of the square roots SmO o i Sum of squares of difference of square roots Elmo foo Sum absolute value of the difference of the log Y Logo nG Log oG Dual Objective Generic calibration can be performed optionally on several objectives As of November 2003 there are three ways to have a secondary objective The background programming logic already allows a much more sophisticated approach and for full multi objective calibration but the current interface has been kept relatively simple on purpose More options are likely to be available in future versions or other Toolkit products Constraint on the total runoff This dual objective penalises the primary objective when the relative total runoff depth difference between the observed and calculated data goes beyond a threshold expressed in percentage Note that since there is in general n
71. ions Nash Sutcliffe criterion Coefficient of efficiency Sum of square errors Root mean square error RMSE Root mean square difference about bias Absolute value of bias Sum of square roots Sum of square of the difference of square root Sum of absolute difference of the log Clicking the mouse on the Plot Curve Response gt gt button runs the model with all other model parameters fixed and varies the selected parameter within the specified bounds and displays the result of the objective function in the graph area 28 rainfall runoff library Issued 3 Nov 03 Using RRL Simulation dialogue The Simulation dialogue as shown in Figure 26 allows the user to run the model for a specified period and output the results The simulation dialogue has four areas output period fields run button output buttons and a graph window Model Input Dates Calibration Sensitivity Simulation Tools Selecta standard plotting option Choose astan Clear Plot Input Data Dates Min 1 01 2010 Max 31 12 2025 Run from 1017 January 2010 Ka to J 017 January 2010 Run Output Desired runoff units moa gt Save Runoff as Save Monthly Runoff as View Recorded Time Series E 5 E Show Graticule Figure 26 Simulation dialogue The output period fields show the earliest and latest dates that are possible to run the model over This is constrain
72. isture stores needs to be made This can be done by assessing the rainfall conditions prior to the start of the model or by selecting a warm up period such that the soil store will be at a known level If the warm up period is wet then all of the soil stores may be full or if the warm up period is dry then the stores may be empty The RRL provides a tool for automatically setting the model warm up period for both calibration and validation Section 3 1 2 The RRL estimates the warm up period by starting the model at different initial condtions and determining where the answers converge If there is no convergence found a warning message is displayed and the warm up period is not set 6 2 4 Setting the periods Having chosen appropriate periods for calibration and validation these periods can be set by entering the appropriate dates in the date fields or by using the slider bars Manual calibration of models The RRL provides many different types of optimisers for calibrating models However manual calibration is also an important aspect of model calibration Manual calibration can be used to investigate how the different parameters change the shape of the simulated hydrograph and also to refine an optimised calibration The RRL provides a manual calibration tab in the calibration dialogue Section 3 1 This tab contains A dynamic update checkbox e List of calibration parameters and An update graph button 6 3 1 Dynamic update
73. l greatest depth of precipitation for a given duration that is physically possible over a particular catchment area based on generalised methods Quickflow The component of streamflow that has travelled through the catchment as interflow or across the surface as overland flow Regulated river The section of river that is downstream of a major storage from which supply of water to irrigators or other users can be regulated or controlled In NSW these storages and rivers are operated by State Water and the regulated rivers are designated by legislation Regulated flow Water that is released from storage to meet downstream requirements Riparian zone The zone adjacent to streams and rivers usually there is some exchange of water and nutrients between this zone and the stream River diversion Water diverted by pump or gravity from a stream Seepage The movement of water downwards through soils or permeable rock This water can originate from a very wide range of sources including all water bodies and most land surfaces Seepage water may percolate far enough to reach groundwater Subcatchment Area of land within a catchment used in specific contexts to distinguish components of a larger catchment see also catchment Transmission loss The flow volume that is lost from a river or stream as water travels downstream It includes seepage to groundwater overbank flow that goes into floodplain
74. le dates for a calibration It also specifies the start end and warm up period for the calibration Verification dates Specifies the minimum and maximum possible dates for a validation lt also specities the start end and warm up period for the validation CalibrationType The calibration type section has exactly the same format and is used to identy the optimisation method used in the calibration Note that most calibration tools are initialised randomly consequently no extra information about the calibration process is stored as it would not allow the reproduction of results Recorders All models developed within TIME can be tagged as State variables that is variables which without having to be considered as primary outputs play a role in the production of the primary outputs The recorders section specities what state of the output variable If the line contains the word True the variable is output Each line also contains the number of signiticant digits specified by the user to be optionally used for display purposes Note that the RRL will always enforce the runoff as being recorded Data scaling The data scaling section specifies the monthly multipliers to apply to the model inputs rainfall and evpotranspiration inputs Comments The comments are stored at the very end of the file Comments should be updated through the user interface The software does not extract any information from it apart from making i
75. led Complex Evolution SCE UA AWBM custom optimisers The objective functions and data transformations that are provided include Nash Sutcliffe criterion Coefficient of efficiency Sum of square errors Root mean square error RMSE Root mean square difference about bias Absolute value of bias Sum of square roots Sum of square of the difference of square root Sum of absolute difference of the log Runoff difference in Flow duration curve Base flow method 2 rainfall runoff library Issued 3 Nov 03 1 3 1 4 1 9 Using RRL 1 2 3 Audience The RRL is intended for a specific audience of users comprising hydrologists in consulting firms government agencies and students learning about rainfall runoff models It is also intended as an adjunct to provide data for other catchment management tools Data requirements 1 3 1 Input data The major inputs to the RRL are as follows e Rainfall a continuous time series of rainfall data that represents the rainfall across the catchment Evapotranspiration a continuous time series of potential evapotranspiration or actual evapotranspiration data that represents the evapotranspiration across the catchment e Flow gaugings daily runoff values for the gauging station that is to be modelled This data are used for model calibration and checking e Catchment area this is used to convert inputs and outputs between flow and depth of runo
76. libration and verification periods however it is good modelling practice that these periods do not overlap Calibration dialogue The calibration dialogue as shown in Figure 20 allows the user to calibrate the rainfall runoff model both automatically and manually The dialogue has five areas 1 2 3 4 5 Boundaries and fixed parameters button Calibration tab Calibrate button Graph window Viewing results buttons 23 Rainfall Runoff Library User Guide Model Input Dates Calibration Sensitivity Simulation Method of Calibration Graph type Choose astandard plot Clear Plot Boundaries and fixed parameters Calculated Vs Observed Generic Custom Manual Nash Sutcliffe Criterion for calibration O Optimisation Method SCE UA Parameters Objective Calculations Primary Objective Nash Sutciiffe Criterion MR pe z Secondary objective y E E Lernens E E Calibrate on monthly values E a T Dynamic Update View o Calibrate D 96 Time Left unknown View Time Series 5 00 10 00 15 00 Calibration Results Observed mm day Data Statistics Tr a Log Show Graticule Figure 20 Calibration dialogue Clicking the mouse on the boundaries and fixed parameter button displays the calibration parameters dialogue as show in Figure 21 The dialogue displays each of the model parameters upper and lower bounds to each parameter an
77. lr nsassem TIME Core Units TimeCore 2C 20Version 3D1 0 1390 33501 2C 20Culture 3Dneutral 2C 20PublicKeyToken 23D1 7delc0433b5448b gt lt _symbol href ref 15 gt lt _name href ref 15 gt factor 1 factor a3 UnitPrefix lt a3 DimensionLess id ref 16 xmlns a3 http schemas microsoft com clr nsassem TIME Core Units TimeCores 2C 20Version 3D1 0 1390 33501 2C 20Culture 3Dneutral 2C 20PublicKeyToken 23D1 7delc0433b5448b lt name id ref 20 gt Dimensionless lt _ name gt Dimension x002B name href ref 20 gt lt a3 DimensionLess gt lt SOAP ENC Array id ref 17 SOAP ENC arrayType xsd anyType 0 lt SOAP ENC Array gt lt SOAP ENC Array id ref 18 SOAP ENC arrayType xsd anyType 0 lt SOAP ENC Array gt lt SOAP ENC Array id ref 19 SOAP ENC arrayType xsd anyType 16 gt lt item href ref 21 gt lt item href ref 22 gt lt SOAP ENC Array gt a3 CompoundUnit x002B PoweredUnit id ref 21 xmlns a3 http schemas microsoft com clr nsassem TIME Core Units TimeCore 2C 20Version 3D1 0 1390 33501 2C 20Culture 3Dneutral 2C 20PublicKeyToken 23D1 7delc0433b5448bp u href ref 23 gt p l p a3 CompoundUnit x002B PoweredUnit a3 CompoundUnit x002B PoweredUnit id ref 22 xmlns a3 http schemas microsoft com clr nsassem TIME Core Units TimeCore 2C 20Version 3D1 0 1390 33501 2C 20Culture 3Dneutral 22C 20PublicKeyToken 3D1 7delc0433b5448b lt _u h
78. lt _label gt lt hasCustomLabel false hasCustomLabel Unit x002B dimensions href ref 16 gt Unit x002B label href ref 34 gt Unit x002B hasCustomLabel gt false lt Unit_x002B_ hasCustomLabel gt a3 SimpleUnit lt a3 SimpleUnit id ref 32 xmlns a3 http schemas microsoft com clr nsassem TIME Core Units TimeCore 2C 20Version 3D1 0 1390 33501 2C 20Culture 3Dneutral 2C 20PublicKeyToken 23D1 7delc0433b5448b gt lt dimensions href ref 16 gt lt _refUnit href ref 35 gt lt _scalingFactor gt 86400 lt _scalingFactor gt lt _offset gt 0 lt _offset gt lt _label id ref 36 gt day lt _label gt lt _hasCustomLabel gt false lt _hasCustomLabel gt Unit x002B dimensions href ref 16 gt Unit x002B label href ref 36 gt Unit x002B hasCustomLabel gt false lt Unit_x002B_ hasCustomLabel gt lt a3 SimpleUnit gt lt a3 RefUnit id ref 33 xmlns a3 http schemas microsoft com clr nsassem TIME Core Units TimeCore 2C 20Version 3D1 0 1390 33501 2C 20Culture 3Dneutral 2C 20PublicKeyToken 23D1 7delc0433b5448b lt baseUnit href ref 37 gt dimensions href ref 38 gt refUnit href ref 33 gt lt _scalingFactor gt 1 lt _scalingFactor gt lt _offset gt 0 lt _offset gt 76 rainfall runoff library Issued 3 Nov 03 Data storage and file formats lt _label id ref 39 gt m lt _label gt lt _hasCustomLabel gt false lt _hasCustomLabel
79. meters that are output are selectable from the Edit Recorded Time Series menu option 3 1 4 RRL Graphs The RRL contains three graph windows in the Calibration Sensitivity and Simulation tabs These graphs are slightly different in each tab but function as described in section 2 5 5 Calibration Graph The calibration graph as shown in Figure 28 allows the user to plot the results of the objective function used for optimised calibration or compare observed and simulated data The graph can be configured such that it updates with each change of parameters or update by clicking on the update button which is discussed above 30 rainfall runoff library Issued 3 Nov 03 Using RRL Selecta standard plotting option Observed and Calculated runaffs Update Objective Clear Plot mm Runoff N A 2 5 Observed and calculated runoffs DN TOT La Show Dates E a lv Show Graticule Figure 28 Calibrate graph The specific options and features that are available on the calibration graph include e Update objective function e Different graph types e log linear y axis e Days date linear log probability x axis e Graticules on off e Clear plot When an optimised calibration is being carried out from the calibration tab the user can choose to display the objective function for successive model runs The graph will be updated for each successive run and is selected by clicking on Update Objective check box All g
80. microsoft com netframework downloads redist aspx You can check to see if you already have the NET Framework installed by clicking Start on your Windows desktop selecting Control Panel and then double clicking the Add or Remove Programs icon When that window appears scroll through the list of applications If you see Microsoft NET Framework 1 1 listed the latest version is already installed and you do not need to install it again Distribution media The RRL is distributed via the Toolkit web site and can be downloaded by 1 Going to the Toolkit web site www toolkit net au Registering as a toolkit user Once a registered toolkit user then going to the users download area and selecting RRL 4 Press the Submit button to commence the download 2 2 1 Licence agreement Acceptance of the licence agreement is part of the installation procedure You must acknowledge that you have read understood and agree to be bound by the RRL software licence agreement to be able to proceed with the installation The licence agreement can be viewed when using the RRL by selecting Help About and selecting the Licence agreement button 2 2 2 Data At present the install shield creates a data directory that contains The rainfall runoff software 2 Documentation User Guide help and training material 3 Example data sets 4 Example project files Directories The RRL installation creates three sub directories within the RRL directory loc
81. mm the evapotranspiration coefficient 0 1 and C the water level of tank 4 6 3 Infiltration The infiltration in each tank is calculated using I C B 3 Where J is the infiltration in mm C the water level of tank x and B the infiltration coefficient tank x 4 6 4 Storage The amount of water in each tank affects the amount of rainfall infiltration evaporation and runoff The storages are calculated from the top to the bottom tank The evaporation is initially deducted from the first storage up to a maximum of the potential rate The remaining potential evapotranspiration is taken from each of the lower tanks until the potential rate is reached or all of the tanks have been evaporated After evaporation has been taken from the tanks rainfall is added to the top tank and based on the revised level runoff and infiltration is estimated This is subsequently deducted from the storage level The next tank subsequently receives the infiltration from the tank above The process continues down through the other tanks 46 rainfall runoff library Issued 3 Nov 03 5 1 5 2 Description of optimisers Description of optimisers Uniform random search This is a very simple optimisation method where by the parameter space for each parameter is divided up into a specified number of intervals between the minimum and maximum bound The optimisation proceeds by randomly selecting from the available options for each parameter t
82. model At present there is only one custom calibration tool and that is for the AWBM model For all other models custom calibration is not available 6 4 1 Custom calibration of AWBM The AWBM custom calibration is a specitic facility written to automatically calibrate the AWBM model To use this facility Select the custom tab in the calibration dialogue 2 In the custom method drop down menu select AWBM automatic calibration currently the only option 3 Select the parameters button to display the optimiser parameters dialogue There are two optimisation parameters that can be set convergence criterion and maximum average capacity Note the default settings are a good place to start 4 Press the calibrate button to start the model The calibratin results can be reviewed by pressing the calibration results button Depending on how well the optimiser is converging the optimisation may need to be adjusted and then recalibrate the model 6 4 2 Generic calibration There are seven optimisation algorithms available Chapter 5 Unitorm random sampling Pattern search Pattern serach multi start Rosenbrock Rosenbrock multi start Genetic algorithm SCE UA N BO oO A WN There are options of having both a primary and secondary objective function There are 8 primary objective functions available Nash Sutcliffe criterion Coefficient of efficiency 2 Sum of square errors 3 Root mean square error RMSE 4 Root mean square
83. modified to take into account that recharge is finite Static water level The level of water in a well that is not being affected by withdrawal of groundwater Unsaturated zone The zone between the land surface and the water table Also known as the zone of aeration or the vadose zone Water table The upper water level of unconfined groundwater where the water pressure is equal to that of the atmosphere and below which the soils or rocks are saturated Waterlogging Waterlogging occurs when the watertable rises into the root zone saturating the soil surface with water from rising groundwater or surface run off collecting in low areas It results in anaerobic absence of free oxygen conditions which reduce plant growth 10 3 Soils Acronym Term Definition Soil hydraulic properties These soil properties control the storage and movement of water salts and chemical through soils They are critical in understanding and predicting runoff recharge and movement of water and salts in catchments Soil landscape An area of land that has recognisable and specifiable topography and soils that is capable of being represented on maps and of being described by concise statements A soil landscape will contain a range of soil types in a defined pattern Soil profile A vertical section of earth from the soil surface to parent material that shows the different soils horizons Soil type A general
84. n capacity becomes infiltration excess runoff 38 rainfall runoff library Issued 3 Nov 03 Description of models infiltration excess runoff SRUN interception store infiltration INF interflow and saturation excess runoff INT groundwater recharge moisture store baseflow BAS groundwater store PET areal potential evapotranspiration input data RUNOFF EXC RAIN JINSC EXC gt 0 INF lesser of COEFF exp SQ x SMS SMSC EXC y SRUN EXC INF INT SUB x SMS SMSC x INF REC CRAK x SMS SMSC x INF INT SMF INF INT REC ET lesser of 10x SMS SMSC PET BAS KxGW Model parameters INSC interception store capacity mm COEFF maximum infiltration loss mm SO infiltration loss exponent SMSC soil moisture store capacity mm SUB constant of proportionality in interflow equation CRAK constant of proportionality in groundwater recharge equation K baseflow linear recession parameter Figure 34 Structure of the SIMHYD rainfall runoff model Moisture that infiltrates is subjected to a soil moisture function that diverts the water to the stream interflow groundwater store recharge and soil moisture store Interflow is first estimated as a linear function of the soil wetness soil moisture level divided by soil moisture capacity The equation used to simulate interflow therefore attempts to mimic both the interflow and saturation excess runoff processes with the soil wetness used to
85. n x002B Length lt a3 BaseUnit id ref 40 xmlns a3 http schemas microsoft com clr nsassem TIME Core Units TimeCore 2C 20Version 3D1 0 1390 33501 2C 20Culture 3Dneutral 2C 20PublicKeyToken 23D1 7delc0433b5448b lt dimensions href ref 16 gt lt _refUnit href ref 35 gt lt _scalingFactor gt 1 lt _scalingFactor gt lt _offset gt 0 lt _offset gt lt _label id ref 45 gt s lt _label gt lt _hasCustomLabel gt false lt _hasCustomLabel gt SimpleUnit x002B dimensions href ref 16 gt SimpleUnit x002B refUnit href ref 35 gt SimpleUnit x002B scalingFactor gt 1 lt SimpleUnit_x002B_ scalingFactor gt 77 Rainfall Runoff Library User Guide SimpleUnit x002B offset gt 0 lt SimpleUnit_x002B_ offset SimpleUnit x002B label href ref 45 gt SimpleUnit x002B hasCustomLabel gt false lt SimpleUnit_x002B_ hasCustomLabel gt Unit x002B dimensions href ref 16 gt Unit x002B label href ref 45 gt Unit x002B hasCustomLabel gt false lt Unit_x002B_ hasCustomLabel gt a3 BaseUnit a3 BaseDimension x002B Time id ref 41 xmlns a3 http schemas microsoft com clr nsassem TIME Core Units TimeCore 2C 20Version 3D1 0 1390 33501 2C 20Culture 3Dneutral 2C 20PublicKeyToken 23D1 7delc0433b5448b name id ref 46 Timec name BaseDimension x002B name href fref 46 XElementaryDimension x002B name href ref 46 gt Dimension x002B nam
86. ng started cae oo oe tO ida 17 TOC 3T Starting the PRL cien oe OR RN EN caps 17 3 1 2 RREMOGDU deett geo edo Ub o o dE iain 17 3 1 3 ARREDIGIOQUES eeepc ette deoa teta dua e on obtu 19 solde RRE GrIODSuo soe REC EO RO E NM euenit Do RE 30 3 2 Gredling d New proletaria e HER I PERIIT R X ACCRUE EROR 34 3 3 Loading an existing BEOIeCE eo st r ER ios 35 4 Description of models ooccoocccoccoccnnnccnnonns 36 4 1 PND e apis e 36 4 2 SACraAME NTO osas aa 37 4 3 o 38 4 4 SMAR a as aan 40 Z1 Walterbalange isc feas 41 AAD gt RODIN Gi As o 42 4 5 SWAT Sus iii aid 44 4 6 TO AER A AAA Tenn eer EON dE 44 4 6 RU da a 45 4 6 2 Evapotranspiration id AAA Cd e AED da 46 4 6 3 AR A dete e t E RE ead ee IRR 46 40 4 SIOIdge o Moe RL toe 46 5 Description of optimisers ooccoocccccnnocnnccnnnonos 47 5 1 U iform random search co ev a Reese HV OUR ERU ERAT SUR HER 47 5 2 Pater Sear c A O A AOA EKE CAY 47 5 3 Malti start pattern search id A ox needed ende edere 48 5 4 Rossrnbrock MEM es Das A seven anas rtu AAE eens 48 5 5 Multi start Rosenbrock search ccceeececeeeeeceececeeeeeeceeeaeeeceeauaeeeeeeanes 51 5 6 Genetic algabHBinitss O a LUKE debe R US DD 51 5 64 Parameter coding s o io alicia 51 66 2 Search method s o eod o er ra t D Veo etica 52 5 7 Shuffled complex evolution ous iced ie i OD rece aa tierra EXER ies 52 6 Model Calibration eese eese 54 6 1 Data Pr
87. nowmelt or 82 raintall runoft library Issued 3 Nov 03 Glossary Acronym Term Definition irrigation see also deep drainage and recharge Influent stream An influent stream is one which gains water from groundwater also known as a gaining stream Lateral throughflow Relatively rapid subsurface flow through cracks pipes macropores in the soil also referred to as lateral subsurface flow and interflow Natural development conditions This term denotes a scenario that comprises conditions in a river system without any water resources development dams water supply infrastructure irrigation development Catchment land use changes that have occurred over the years such as land clearing are ignored in this definition This term is relevant to hydrologic modelling for water management purposes and Water Sharing Plans Overland flow Surface runoff which is caused either because the underlying soil is saturated and cannot accommodate any more water or because the intensity of rainfall is greater than the soil s capacity to infiltrate it Precipitation Rain snow hail sleet dew Probable Maximum Flood PMF The flood resulting from PMP and where applicable snow melt coupled with the worst flood producing catchment conditions that can be realistically expected in the prevailing meteorological conditions Probable Maximum Precipitation PMP The theoretica
88. nt that an appropriate calibration period be selected The model should be preferably calibrated in a period that has both wet and dry extremes and has an average annual flow similar to the average annual flow for the whole period of record The RRL provides two buttons in the date dialogue that will show the wettest and driest annual flow periods Section 3 1 2 When selecting a continuous calibration period it is not always possible to get both wet and dry extremes in the calibration period In most cases it is more important to include the dry extreme as this will better set the size and discharge rates of the soil stores If possible the validation and calibration periods should be of similar length However to include sufficient climatic variability in the calibration period it may be possible to have the validation period only cover one third of the period of record There are several other issues that should be taken into consideration when selecting a calibration period The intended use of the model Is the model to be used for yield analysis water allocation or flood prediction Over the period of flow record has there been any change to the methods of measuring flow height recorders and rating curves Has there been any significant change in land use over the period of flow record If the model is to be used for yield analysis or water allocation then selecting a calibration period that represents average or dry conditions
89. ntium class processor Operating The RRL is supported on the following platforms System Microsoft Windows Server 2003 NET Framework 1 1 is installed as part of the operating system Windows XP Professional Windows XP Home Edition Windows 2000 Windows Millennium Edition Windows Me Windows 98 Microsoft Windows NT 4 0 Service Pack 6a The RRL cannot be installed on 64 bit computers Memory 128 megabytes MB of RAM 256 MB recommended Hard Disk 120 MB of hard disk space required 40 MB additional hard disk space required for installation 160 MB total Display 800 x 600 or higher resolution display with 256 colours Input Device Microsoft mouse or compatible pointing device Other e Microsoft Internet Explorer 5 01 or later is required e The net framework 1 1 redistributable or later is required Table 2 System requirements The person installing the software must have local administrator access on the computer Rainfall Runoff Library User Guide 2 2 2 3 2 1 1 Framework The NET Framework is a component of the Microsoft Windows operating system used to build and run Windows based applications The RRL is built using the NET Framework and consequently requires that the NET Framework redistributable be installed prior to running the RRL The NET Framework Redistributable will install the NET Framework onto your machine and is downloadable from www msdn
90. o certainty that the model is structurally capable of respecting this constraint the result of the calibration may still give results beyond the specified threshold Dual calibration on the flow duration curve This option uses the same objective as the one selected as the primary objective with the same parameters e g lambda Box Cox but applies it to a sample of the duration curve 100 values are taken from the curve for each percentile Since there is a loss of information i e the dates between the time series and the duration curve it is wise to keep an intermediate value for the weighing of the two objectives this weighing parameter can be accessed via the Parameter button of the second objective 62 rainfall runoff library Issued 3 Nov 03 6 5 Model Calibration Dual calibration on the baseflow This option uses the same objective as the one selected as the primary objective but is applied to the baseflow output of the model The baseflow calibrated against is derived from the observed runoff depth by the use of a baseflow filter Grayson 1996 There are two parameters controlling this filter plus a weighing parameter for the weighing of the two objectives accessible via the Parameter button of the second objective The calculated Baseflow is either taken from the appropriate model output or filtered from the calculated runoff if the model does not define any Baseflow as an output Selection of the objective func
91. o values are separated by a comma An example of this format is show in Figure 45 1 01 1900 7 2 01 1900 11 3 01 1900 19 4 01 1900 5 5 01 1900 0 6 01 1900 19 Figure 45 Example of CSV daily time series format 8 2 5 IQQM daily time series format iqqm An IQQM daily time series format file is an ASCII text file that contains daily time series data The file has a five line header followed by annual tables of daily data The five lines of information contained within the header are The title line which is a 40 character string of information detailing how the file was created It is suffixed with a date and time of creation 2 The site name which is a 40 character string specitying the site for which the data applies 3 The type which is a 15 character string specifying the data type within the file eg precipitation evaporation or gauged flow 69 Rainfall Runoff Library User Guide 4 The unit line which is a 10 character string specifying the units of data eg mm mm 0 1 ML d and 5 The data line which specifies the time span of the time series data and the time interval of data stored within the file In daily IQQM format files the daily data is grouped in tables of yearly data with 31 columns representing each day in a month and 12 rows for each month of the year At the end of each row of daily data is a monthly total and at the end of each table is a yearly total These tables are repeated for
92. odel be checked by assessing How representative the rainfall period is i e is it wet dry or average How close are the extremes of wet and dry to the extremes in the long term record Are there periods when the soil stores are fully saturated and fully dry Are the proportions of surface flow interflow and base flow appropriate given the catchment size catchment shape soil type and land use How representative the rainfall period is can be checked by plotting up the annual rainfall volumes of each year in an exceedance plot The extremes of rainfall during the period of flow record can be checked to see it they lie within the bottom and top thirds of the the annual rainfall excedence The mean annual rainfall for the calibration period can be checked against the mean annual rainfall for the entire period of record If the dry extreme is above the lower third of rainfall and or the mean of the calibration period is one standard deviation above the mean of the entire period of record then it is likely the calibration will have a bias toward wet periods This will mean that the model wil perform well during wet periods but is likely to overestimate flows during dry periods If the wet extreme is below the higher third of rainfall and or the mean of the calibration period is one standard deviation belowthe mean of the entire period of record then it is likely the calibration will have a bias toward dry periods This will mean that the mo
93. omponent A schematic diagram of the SMAR model is shown in Error Reference source not found The model utilises time series of rainfall and pan evaporation data to simulate stream flow at the catchment outlet The model is calibrated against observed daily stream flow The water balance component divides the soil column into horizontal layers which contain a prescribed amount of water usually 25 mm at their field capacities Evaporation from soil layers is treated in a way that reduces the soil moisture storage in an exponential manner from a given potential evapotranspiration demand The routing component transforms the surtace run off generated from the water balance component to the catchment outlet by a gamma function model form Nash 1960 a parametric solution of the differential routing equation in a single input single output system The generated groundwater run off is routed through a single linear reservoir and provides the groundwater contribution to the stream at the catchment outlet The SMAR model contains five water balance parameters and four routing parameters The routed surface and the groundwater run off from the SMAR model are used in the salt mobilisation and wash off component of CATSALT to estimate the associated salt load 40 rainfall runoff library Issued 3 Nov 03 Description of models Evaporation B Conversion to Rainfall potential rate R T Excess rainfall A n AUX Direct run off
94. on the y axis 81 30 18902 2105 1907 31 08 1912 3112 1917 Table 3 TIME fields 2 5 5 Graph Windows TIME provides a module to plot specific information as graphs Figure 4 shows an example of the basic form of graphs provided in TIME The graph is displayed in a window and contains Title text X and Y axes with optional labels Graphed data points Optional graticules Optional legend Nash Sutclifte Criterion f Uzk 1 00 0 99 0 99 0 98 0 98 0 97 0 97 0 96 0 05 0 10 0 15 0 20 0 25 0 30 0 35 0 40 0 45 Figure 4 Graph window By right clicking the mouse in the graph window the graph menu popup is displayed as shown in Figure 5 Pan Zoom In Copy Graph Properties Figure 5 Graph popup 12 rainfall runoff library Issued 3 Nov 03 Using RRL Four menu options are provided in the graph popup e Pan e Zoomin e Copy Graph e Properties The graph screen operates in two control modes pan or zoom in The current graph control mode can be determined by viewing the graph menu popup and is signified by a dot to left of the menu item Alternatively the graph control mode can be identified by the cursor shape a cross indicates Zoom In and a cross with arrows indicates Pan Control mode Figure 5 shows that Zoom In is the current control mode To change the graph control model click the mouse on the appropriate text Pan When the graph window is in the Pan Control mode the graph may be paned in any di
95. peed up the calibration process The rainfall runoff models require four important data sets 1 Catchment characteristics 2 Rainfall 3 Evapotranspiration and 4 Flow 6 1 1 Catchment characteristics Generally the only catchment characteristic required by lumped rainfall runoff models is the catchment area However in some cases models e g SWAT need to know slope land use soil profile soil depth and hydraulic conductivity The models operate in mm and to convert the model output from runoff depth to runoff volume catchment area is required The catchment area is usually an easy parameter to obtain but should be used with caution The area is dependent on the scale of maps or DEMs that it was derived from and in flatter areas there can be large uncertainty with regard to where catchment boundaries are A small error in cathment area can cause a large error in the estimated volume that runs off the catchment Although slope land use soil profile soil depth and hydraulic conductivity may not be used by a model this information is also worth considereing The type of land use will influence surface runoff characteristics evpotranspiration rates and interception losses The soil characteristics will influence the size of soil stores and seepage rates This sort of information 54 rainfall runoff library Issued 3 Nov 03 Model Calibration is invaluable for setting realistic bounds on model parameters as well as sanity checking
96. rage rainfall on the catchment each month and then disaggregate this data to daily data with selected Theissen weighted rainfall sites 6 1 3 Evapotranspiration data There are many different methods of estimating evapotranspiration A few of the common methods are listed below e Evaporation pan Class A sunken tank sunken tank with bird gard Lysimeter Priestely Taylor equation e Morton equation e Penman equation and Penman Monteith equation The important issue with evaporation data is that all of these different methods will provide a different estimate of evaporation It is important that whatever source is used that it is consistent with what the model requires There are typically two different types of evapotranspiration data required by models potential evapotranspiration Ej and actual evapotranspiration E There are factors that can be applied to each of these data sources to convert to the appropriate type of evapotranspiration The RRL has a data scaling dialogue that allows annual or monthly factors to be applied to evaporation data 6 1 4 Flow data The flow data is what the rainfall runoff model is calibrated against and in calibrating against the flow data the assumption is made that this data has no errors This is not the case and consequently care should be taken to ensure that the flow data is of good quality There are several things that need to be considered How height data is collected
97. range ai bi The parameter range is discretised into 2l points and the discretisation interval is b a Ax 1 2 1 For example when 7 the mapping is shown in Table 7 Connecting the codings of all parameters forms the coding for each point in the space to be searched Table 7 Example of parameter coding for l 7 Binary Code Integer Value Parameter Value 0000000 0 Ai 0000001 1 Ai dxi 0000010 2 Ai 2dxi 1111110 126 Ai 126dxi 1111111 127 Ai 127dxi bi 3 Rainfall Runoff Library User Guide 5 7 5 6 2 Search method The search is carried out in the following steps 1 Locate m points randomly in the search space m 100 can be used 2 Find the objective function value at each point 3 Rank the objective function values in descending order 4 Assign a probability value pj to each point 1 2 m giving higher probability to points with a lower function value The worst point after ranking is j 1 and its probability value pi will be the smallest The best point is j m and its probability value pm will be the largest The probability values for other points are linearly interpolated as pepe B A m 1 2 5 The summation of probability values for all points should be equal to unity The average of probability values for all points is then 1 m A value C m can be assigned to pm so that the probability value for the best point is C time the average where C2 1 The correspond
98. ranspired by plants see also infiltration and recharge in Groundwater list End of system flow Streamflow at the end of a given river system eg Murrumbidgee River at Balranald En route storage A water storage weir or lake located either instream or off stream in the middle or lower part of a river system eg Hay Weir on the Murrumbidgee Flow duration curve A graphical representation of a ranking of all the flows in a given period from the lowest to the highest where the rank is the percentage of time the flow value is equalled or exceeded These curves may be derived for flows in any time interval such as daily flows monthly flows or annual flows Flow percentile The percentage of time for which a given flow is equalled or exceeded see also percentiles in Salinity list Flow routing The change in magnitude speed and shape height and duration of a flow hydrograph as it moves downstream due to channel and floodplain storage and frictional effects GL gigalitre One thousand million litres also equal to 1 million cubic metres ydrologic cycle The circulation of water from the oceans through the atmosphere to the land and ultimately back to the ocean ydrograph The trace which describes the change in stream flow or water level over time at a given location in a river system nfiltration The process by which water soaks into the soil from rainfall s
99. raph fields except the Update Objective and Show Graticule check boxes and Clear Plot button will be greyed until the model is run either as an optimised calibration or manual calibration Once the model is run the graph type fields will be activated and the required graph type can be displayed by selecting from the list of graph types in the Select a standard plotting option drop down list The graph plotting options are e Daily Scatter e Monthly scatter e Daily time series calibration and validation periods e Daily difference calibration and validation periods e Daily flow duration calibration and validation periods The graph axis options are dependent upon the graph type that is selected and are listed in Table 5 Graph type x date x log x probability y log Daily Scatter No Yes No Yes Monthly scatter No Yes No Yes 3 Rainfall Runoff Library User Guide Graph type x date x log x probability y log Daily time series Yes No No Yes Daily difference Yes No No Yes Flow duration Yes No Yes Yes Table 5 Calibration graph axis options The x date when checked displays dates on the x axis and when not checked it is days since start date The x log when checked plots a log x axis and when not checked it is a linear x axis The x probability when checked plots a probability x axis when not checked it is a 0 100 sale x axis The y log when checked plots a log y axis and when not checked it is a linear y axis The curren
100. rection by holding down the left mouse key moving the mouse and releasing the mouse key The graph will then be redisplayed As an example to scroll the graph to the left position the mouse to the right of the graph hold the left mouse button down move the mouse to the left of the graph and then release the mouse button Zoom In When the graph window is in the Zoom In control mode the graph may be enlarged in a particular area by positioning the mouse near the edge of the area to be enlarged holding the left mouse button down a window will appear then move the mouse to the other edge of the area to be enlarged and release the mouse button The graph will be redisplayed for the zoomed in area Copy Graph Clicking on the Copy Graph menu item will copy the contents of the graph to the clipboard The graph may then be placed into other Windows applications Properties Clicking on the Properties menu item displays the graph properties tabbed dialogue as shown in Figure 6 This dialogue allows the user to adjust the text colours and fonts of the title axis labels axis and graphed parameters The decorators that are available are dependent upon what is currently displayed in the graph but typically include e View title e X Axis label e Y Axis label e Axis e Graphed variable 1 etc 13 Rainfall Runoff Library User Guide Ma Property Editor Decorators font Microsoft Sa text Calculated Vs textColor _ Whi
101. ref ref 24 gt lt _p gt 1 lt _p gt a3 CompoundUnit x002B PoweredUnit gt lt a3 CompoundUnit id ref 23 xmlns a3 http schemas microsoft com clr nsassem TIME Core Units TimeCore 2C520Version 3D1 0 1390 3350152C 20Culture 3Dneutral 2C 20PublicKeyToken 3D1 7delc0433b5448b gt lt units href ref 25 gt lt prefix href ref 14 gt lt _ label href ref 15 gt lt _hasCustomLabel gt false lt _hasCustomLabel gt lt Unit_x002B_ dimensions href ref 16 gt Unit x002B label href ref 15 gt Unit x002B hasCustomLabel gt false lt Unit_x002B_ hasCustomLabel gt lt a3 CompoundUnit gt lt a3 CompoundUnit id ref 24 xmlns a3 http schemas microsoft com clr nsassem TIME Core Units TimeCore 2C 20Version 3D1 0 1390 33501 2C 20Culture 3Dneutral 2C 20PublicKeyToken 23D1 7delc0433b5448b gt lt units href ref 26 gt lt prefix href ref 14 gt 75 Rainfall Runoff Library User Guide lt _label href ref 15 gt lt _hasCustomLabel gt false lt _hasCustomLabel gt lt Unit_x002B_ dimensions href ref 16 gt Unit x002B label href ref 15 gt Unit x002B hasCustomLabel gt false lt Unit_x002B_ hasCustomLabel gt lt a3 CompoundUnit gt lt a2 ArrayList id ref 25 xmlns a2 http schemas microsoft com clr ns System Collections gt lt _items href ref 27 gt size 1 size lt version 1 version lt a2 ArrayList gt lt a2 ArrayList id
102. reflect parts of the catchment that are saturated from which saturation excess runoff can occur Groundwater recharge is then estimated also as a linear function of the soil wetness The remaining moisture flows into the soil moisture store Evapotranspiration from the soil moisture store is estimated as a linear function of the soil wetness but cannot exceed the atmospherically controlled rate of areal potential evapotranspiration The soil moisture store has a finite capacity and overflows into the groundwater store Base flow from the groundwater store is simulated as a linear recession from the store The model therefore estimates runoff generation from three sources infiltration excess runoff interflow and saturation excess runoff and base flow 39 Rainfall Runoff Library User Guide 4 4 SMAR The soil moisture and accounting model SMAR is a lumped conceptual rainfall run off water balance model with soil moisture as a central theme O Connell et al 1970 Kachroo 1992 Tuteja and Cunnane 1999 The model provides daily estimates of surface run off groundwater discharge evapotranspiration and leakage from the soil profile for the catchment as a whole The surface run off component comprises overland flow saturation excess run off and saturated through flow from perched groundwater conditions with a quick response time The SMAR model consists of two components in sequence a water balance component and a routing c
103. rently specified graph to be updated This can be a time consuming process if the calibration period is long in which case the user can turn this option off and use the Update Graph button as required The model parameters that are displayed are dependent upon the model that is selected and consequently the list of parameters will vary for each rainfall runoff model The user can simply type in the new value or use the spinner for the parameter to be changed Note range checking is performed based on the boundaries specified in the calibration parameters dialogue that is displayed by clicking on the boundaries and fixed parameter button The Update Graph button allows the user to control when the rainfall runoff model is run and the graph is updated for the current set of model parameters Clicking on the Calibrate button will cause the program to calibrate the selected rainfall model using the specified optimisation method subject to the selected objected function and data transformation Progress of the optimisation is displayed at the bottom of the calibrate tab dialogue If the Update Objective check box is checked then the Objective function values will be displayed in the graph window and progressively update for each model run The graph window has a selection of options for displaying observed and simulated flows as well as the option to display the value of the objective function This is discussed in more detail in section 2 5 5 The view
104. rve the elevation of the water table or the potentiometric surface Also known as a piezometer Overburden The loose soil silt sand gravel or other unconsolidated material overlying bedrock Also known as regolith Perched aquifer watertable A watertable above the main watertable level where impermeable soil or rock prevents the water from percolating through it to the main groundwater body Permeability The property or capacity of a porous rock sediment or soil for transmitting water Sand for example is said to have high permeability Piezometer see Observation well bore Production bore pumping well A bore equipped with a pump to extract groundwater from an aquifer system Recharge The component of deep drainage that reaches the groundwater table see also deep drainage and infiltration both in Surface Water list Regolith see Overburden Safe yield The volume of water that can be taken from an aquifer at any given 84 rainfall runoff library Issued 3 Nov 03 Glossary Acronym Term Definition location regardless of impacts which is largely dependent on the type of bore and capacity of pump that can be installed This definition assumes the recharge of the aquifer is limitless which of course it is not in reality so it can encourage mining of the resource and lead to environmental damage Increasingly the usage of the term safe yield is being
105. s The complexes are partioned such that the first complex contains every p k 1 1 ranked point the second complex contains every p k 1 2 ranked point and so on where k Evolve each complex according to the competitive complex evolution CCE algorithm as detailed below Shuffle complexes by combining the points in the evolved complexes into a single sample population sort the sample population in order of increasing objective function value shuffle the sample population into p complexes according to the procedure specitied in Step 3 Check for convergence by checking if any of the pre specitied convergence criteria are met stop otherwise continue Check the reduction in the number of complexes if the minimum number of complexes required in the population Pin is less than p remove the complex with the lowest ranked points set p p and s pm return to Step 4 If pmin p return to Step 4 53 Rainfall Runoff Library User Guide 6 1 Model Calibration This Chapter discusses the things that need to be considered when calibration a rainfall runoff model The chapter covers 1 Data preparation 2 Calibration and validation periods 3 Manual calibration of models 4 Automatic calibration of models 5 Sensitivity of parameters Data preparation The most important step in calibrating a raintall runoff model is data preparation Time spent in ensuring that the best possible data set is used will greatly s
106. s a version of the User Manual converted into a windows help file This tile can be viewed by double clicking on the file The file is also used by the RRL to display help via the help menu 8 1 5 Documentation directory The documentation directory has two sub directories one which contains a PDF version of the User Manual and the other contains training material File formats 8 2 1 AWBM daily time series format An AWBM daily time series format file is an ASCII text file that contains daily time series data There is no header line in the file The data is organized in rows of one month of data seperated by the spaces The first entry in a row is the number of days in the month This is followed by data values for each day in the month The data values are followed by the year and month of the data An example of this format is show in Figure 42 31000000 00001 20000 0 08 0 8 15 28 0 88 0 08 0 0 0 2001 1 28000000 1 6800000 0 24 0 56 3 28 0 88 0 4 0 1 28 3 12 2001 2 31 0 08 0 0 0 0 0 08 0 8 0 2 48 0 8 3 28 1 28 0 08 0 24 0 0 88 2001 3 Figure 42 Example ofAWBM daily time series format 8 2 2 Chiew and Siriwadena daily rainfall time series format rain An Chiew and Siriwadena daily rainfall time series format file is an ASCII text file that contains daily time series rainfall data There is no header line in the file The dota is organized as date followed by data value The date is in year 4 characters month 2
107. s likely that a minor version upgrade would require adjustments to documentation data examples and training material Bug Fix Version increment This indicator is incremented after minor corrections to code and is not associated with any changes to documentation data examples or training material In some cases users will be notitied of the implications of particular bug fixes 2 5 2 Menus TIME uses standard Microsoft Windows menus as shown in Figure 1 The features in the menu are the e Top level menu items e Menu drop down lists e Popup lists Top level menu Top level menu items may be selected by clicking the mouse on the required text or by using hot keys to select the menu option Note on some versions of word hot keys are identified by underlining the relevant character in the menu item string The hot key is essentially the alt key in combination with the underlined letter Drop down menu list If a selected top level menu item contains a drop down list the list will be displayed as shown in Figure 2 Rainfall Runoff Library User Guide Content Search Index Toolkit Web Site Technical Support Software Update Register RRL About RRL Figure 2 Drop down menu list Drop down list items may be selected by clicking the mouse on the menu line of text hot keys or ctrl keys If ctrl keys are provided these keys will be listed to the right of the menu item e g F1 after Content in Figure 2 Popup lis
108. s will be cleared 21 Rainfall Runoff Library User Guide Dates dialogue The dates dialogues as shown in Figure 18 allows the user to 1 Locate the wettest and driest annual flow periods within the data 2 Specify a calibration period 3 Specify and optional verification period 4 Find appropriate warm up periods for calibration and verification periods Model Input Dates Calibration Sensitivity Simulation Tools update gt gt gt Lowest runoff Highest runoff Find Calibration Warmup Dates of Calibration Stat 017 January 2010 Warmup 11 April 2010 a a L OZ End o3 February 2020 AAA Al I Verification lv Perform verification ee Start 18 February 12020 L i i D 1 D i D 1 n 1 Warmup 197 July 2020 Alla re va a ath ne A oie hin Save abe 6 End 31 December 2025 ales enS a Snn J Figure 18 Dates dialogue Upon entering this dialogue no time series information is displayed To display the time series data click the mouse on the update button The time series flow rainfall and evaporation will then be displayed Note all other buttons and date fields will be greyed until this button is selected Clicking the mouse on show driest year or show west year buttons will display the annual driest and west periods in the respective suffixed transparent colours The displayed periods can be toggled by clicking on the respective buttons Th
109. t available to read or edit Rainfall Runoff Library Job File version 20020726 Original File Name C src TIME Applications RainfallRunoff Projects AWBMJardineRiver jobf Created on 7 3 2003 11 19 43 AM IMPORTANT you should not modify anything before the section starting with an Tot otherwise it is likely that the file will not be readable Please include any comment after the Model TIME Models AWBM AWBM Location Site Area Area Units Input evapotranspiration bin Debug Data QLD JardineRiver JardineRiver evapotranspiration daily tts runoff bin Debug Data QLD JardineRiver JardineRiver runoff daily tts 72 rainfall runoff library Issued 3 Nov 03 Data storage and file formats rainfall bin Debug Data QLD JardineRiver JardineRiver rainfall daily tts NullValue 9999 Parameter parameter name value Min Max is fixed KBase 0 973797280616433 0 1 False KSurf 0 157963349455041 0 1 False A2 0 437898762571088 0 0 5 False A1 0 480340269624826 0 1 False BFI 0 991380037523361 0 1 False Cl 122 021687079944 0 200 False C3 7573 96082041227 0 10000 False C2 78 4608507255025 0 3000 False CalibrationDates calibrationMinDate 1 01 1902 12 00 00 AM calibrationStartDate 1 01 1902 12 00 00 AM calibrationWarmupToDate 3 01 1903 12 00 00 AM calibrationEndDate 28 03 1905 12 00 00 AM calibrationMaxDate 31 12 1917 12 00 00 AM validationMinDate 1
110. t graph can be cleared by clicking the mouse on the clear plot button Sensitivity Graph The sensitivity graph as shown in allows the user to plot the sensitivity of the calibration of a particular model parameter with respect to a specified objective function The x axis is the number of iterations and the y axis is the value of the objective function The graph is updated by clicking on the plot curve response button which is discussed above Clear Plot Nash Sutcliffe Criterion f Lztwm 100 00 150 00 200 00 250 00 300 00 350 00 Figure 29 Sensitivity graph The current graph can be cleared by clicking the mouse on the clear plot button 32 rainfall runoff library Issued 3 Nov 03 Using RRL Simulation Graph The calibration graph as shown in Figure 30 allows the user to plot the simulated results of the model The graph can be configured to display the simulated results as daily time series monthly time series or daily flow duration Selecta standard plotting option Calculated Runoff v Clear Plot E e Calculated Runoff 0 Hr oh PR ea er PEA RA AAA AE iu A 1153 6153 16153 21153 26153 31153 36153 Dates Days Show Dates E E Show Graticule Figure 30 Simulation graph The specific options and features that are available on the simulation graph include e Different graph types e Log linear y axis e Days date linear log probability x axis e Clear plot All grap
111. te Visible True ScatterPlot Figure 6 Graph properties dialogue Each of the decorators allows the user to adjust different properties dependent up on the decorator Table 4 lists the properties that can be changed on each tab Tab Text Font Text Text postion Line colour Axis colour Title Yes Yes Yes No No No x axis Yes Yes Yes No No No y axis Yes Yes Yes No No No Axis Yes Yes Yes Yes Yes Yes Variable No No No No Yes No Table 4 Configurable graph properties Text fields can be moditied by clicking on and editing the text string Clicking the mouse on the font field button will display a standard Microsoft Windows font dialogue The required font font style and size can be specified in this dialogue Clicking the mouse on the text or line colour field will display a colour selection tab dialogue as shown in Figure 7 This tab dialogue provides three colour selection tabs custom web and system The custom tab displays the standard Microsoft Windows colour selector The web and system tabs allow selection from an existing set of colours Custam Web System NI NMN NI ar ma ALA VA EM ona LLO BES EERE BEEBE EEE LL TELS DS DL Figure 7 Colour picker Clicking the mouse on the text position field will display a drop down list box that allows the user to position the axis label to the top left bottom or right of the graph Top and bottom are valid options for the x axis while left and right
112. the number of years of time series data An example of a daily IQQM format file is shown in Figure 46 The first line of each table specifes the year related to the data within the table This year may be optionally followed by factor Where the number following the factor is a factor that is applies to the whole of the table If this factor does not exist then no factor is applied to the table The time series data values may be suffixed by a special character which modifies the time series data value The special characters and their function is described below 1 Indicates the time series value is to be multiplied by 1000 2 e Indicates the time series value is estimated 3 E Indicates the time series value is estimated and is to be multiplied by 1000 4 n Indicates the time series value is negative Note negative numbers not followed by a n are assumed to be missing 5 N Indicates the time series value is negative and is to be multiplied by 1000 Note negative numbers not followed by a N are assumed to be missing and 6 Indicates that the time series value is missing Note typically missing values are flagged as 12 A negative value not suffixed with a n or N is also considered as missing e g 1 is considered as a missing value Title Winding Rivere at Middle Gauge Date 30 07 2003 Time 17 47 24 66 Site Winding River Type Flow Units ML d Date 01 01 19
113. this simple conceptualisation the behaviour of the tank model is not so simple The behaviour of the model is strongly influenced by the content of each of the stores Under the same rainfall and different storage volumes the runoff generated is significantly different The tank model is applied to analyse daily discharge from daily precipitation and evaporation inputs The concept of initial loss of precipitation is not necessary because its effect is included in the non linear structure of the tank model 44 rainfall runoff library Issued 3 Nov 03 Description of models TANK MODEL 4 St TANK ETA a11 ee 1 a12 Hit i H12 M 2 nd TANK Esa 2 a21 gt H21 2 3 rd TANK ETA 3 E G31 gt mio La 4 th TANK oan 4 a41 gt Figure 36 Structure of the TANK model 4 6 1 Runoff The total runoff is calculated as the sum of the runoffs from each of the tanks The runoff from each tank is calculated as 45 Rainfall Runoff Library User Guide nx 4 q Y Y C H a 1 x l y Where q is the runoff depth in mm C the water level of tank x H the outlet height and a is runoff coefficient for the respective tank outlet Note if the water level is below the outlet no discharge occurs 4 6 2 Evapotranspiration The evapotranspiration is calculated using Beken s 1979 equation 4 ETA M 1 exp aC 2 x l Where ETA is the evapotranspiration in
114. tion The selection of the objective function will give the calibration a bias toward the range of flows that the objective function determines as most significant The intended use of the model should be taken into consideration when selecting the objective function The selected object should give a bias towards the flow characteristics that are of greatest importance such as overall volume monthly volume surface runoff and base flow Running the optimiser Once an optimiser has been chosen and a combination of objective functions selected the optimised calibration can begin The user can check the dynamic update check box to display the current value of the objective function in the graph window This facility does add some overhead to the model run time and for long calibration periods with slow models this option can be turned off to speed up the optimisation process To start the optimisation the calibration button should be clicked A progress percentage optimisation time and progress bar will show the progress of the optimisation Note the optimisation can be stopped at any time by clicking on the stop button When completed the graph can be viewed by selecting a graph display option The progress of the optimiser can be viewed by clicking on the calibration results button The stats of the calibration can be viewed by clicking on the data statistics button The view button recalls the latest graph of objective functions following
115. to the overall change achieved by the stage just completed The motive for this rotation is that the overall direction just obtained is likely to be promising Note that in the case of Error Reference source not found Figure 39 this happens not to be true because we optimised into the narrow valley However the second vector of the new base is promising 49 Rainfall Runoff Library User Guide 0 75 0 7 0 65 0 6 0 55 0 5 0 45 0 4 0 35 1 1 1 0 9 0 8 0 7 0 6 0 5 Figure 39 First orthonormalisation of the base The second stage proceeds in a similar manner to the first one Note in Figure 40 that within two stages the new base does have a vector almost collinear with the steepest slope in the narrow valley 1 0 9 0 8 0 7 0 6 0 5 0 8 0 75 0 7 0 65 0 6 0 55 0 5 0 45 0 4 0 35 0 3 Figure 40 Second orthonormalisation of the base 50 rainfall runoff library Issued 3 Nov 03 5 5 5 6 Description of optimisers Multi start Rosenbrock search The initial sampling of the parameter space provides the potential for locating the global optimum without being biased by pre specified starting points This method works by dividing the parameter values into a specified number of increments between the specified bounds For each of these possible starting points a Rosenbrock search is carried out The best optimum of the
116. ts If a selected drop down menu item has an associated popup list signified by C on the right hand side of the menu item The popup list will be displayed as shown in Figure 3 when the cursor is moved over the menu item line of text Window Help MODELUNGTOOLKIT Calculator Find Ctrl F 3 S Configuration Colour Setup Display Fonts Graph Colours Graph Fonts Figure 3 Popup list Popup menu items may be selected by clicking the cursor on the menu line of text hot keys and ctrl keys 2 5 3 Toolbar buttons There is no toolbar in the RRL window 2 5 4 Fields TIME uses standard Microsoft Windows and purpose built fields that are assembled together to create user interfaces for products The components used to create these interfaces include combinations of dialogues tabs TIME specific display windows and fields The common TIME fields and control features are described in Table 3 String fields The string field can either be edited or Site replaced To edit the field click near the text you want to enter then use the standard edit keys del back space and left right arrows to edit the text Alternatively double clicking on the text will highlight the text and then any subsequent typing will overwrite the entire text Integer fields There are two types of integer fields i e 4 either with or without spinners If a spinner is provided by clicking on the up arrow the number will increase by one and clicking on
117. us to equation 1 as in equation 6 l 1l T h t ld 6 dra ra K rech The recharge O t and the discharge Of f can be time averaged to mm d in an analogous manner as in equations 2 and 3 SWAT core The model works by solving the following water balance equation SW Sw YlR O Es Wee Q 1 where SW is the final soil water content mm H50 SW is the initial soil water content mm H20 Qis the amount of surface runoff on day i mm H 0 E is the amount of evapotranspiration on day i mm H20 Wesp is the amount of percolation and bypass flow exiting the soil profile bottom on day i mm H 0 and Q is the amount of return flow on day i mm H 0 gw The swat code is not discussed in this document A full explanation of the SWAT model is supplied in the SWAT reference manuals Tank The tank model is a very simple model composed of four tanks laid vertically in series as shown in Figure 36 Precipitation is put into the top tank and evaporation is subtracted sequentially from the top tank downwards As each tank is emptied the evaporation shortfall is taken from the next tank down until all tanks are empty The outputs from the side outlets are the calculated runoffs The output from the top tank is considered as surface runoff output from the second tank as intermediate runoff from the third tank as sub base runoff and output trom the fourth tank as base flow Despite
118. ved in the RRL subdirectories will also be deleted if this action is taken 16 rainfall runoff library Issued 3 Nov 03 3 3 1 Using RRL Using RRL Getting started 3 1 1 Starting the RRL To run the rainfall runoff library click on the RRL icon on the desktop or select start programs Toolkit Rainfall Runoff Library The RRL introduction splash screen will appear for approximately 5 seconds or until a mouse button or key is pressed The RRL menu will then appear 3 1 2 RRL Menu The RRL top level menu is a standard Microsoft Windows menu as shown in Figure 9 The menu options and associated hot keys are e File alt f e Edit alt e e View alt v e Run alt r a e Tools alt t e Window alt w e Help alt h gt Rainfall Runoff Library MODELUNGTOOLKIT a Figure 9 RRL top level menu The right hand side of the menu contains the Catchment Modelling Toolkit logo File drop down menu The File drop down menu is shown in Figure 10 This menu allows users to e Create open view print save save as and close RRL project files 17 Rainfall Runoff Library User Guide e Setup the printer e Change the working directory e Select from up to 4 previously opened files e Exit the RRL Edit View Tools Help New Project Open Project 1 Jinfall Runoff Library Beta Projects AVWVBMJardineRiver jobf 2 Jkit Rainfall Runoff Library Beta Projects AWBM_Allyn job
119. w lateral drainage evapotranspiration ET and vertical drainage percolation The SAC SMA model is very often represented graphically as illustrated in Figure 33below 37 Rainfall Runoff Library User Guide Sacramento Soil Moisture Accounting Model ep 29 rorspirotion ET y nit stios F j AP tty East Pto Lower N Zone 5 waaka Ourise iA Riverside Technology inc Figure 33 Structure of the Sacramento rainfall runoff model 4 3 Simhyd SIMHYD is a daily conceptual rainfall runoff model that estimates daily stream flow from daily rainfall and areal potential evapotranspiration data Data formats are described in Chapter 1 and Chapter 8 respectively SIMHYD is a simplified version of the daily conceptual rainfall runoff model HYDROLOG that was developed in 1972 see Porter 1972 and Porter and McMahon 1975 1976 and the more recent MODHYDROLOG Chiew and McMahon 1991 The SIMHYD model has 7 parameters as compared to the 17 parameters required for HYDROLOG and the 19 for MODHYDROLOG The structure of the simple lumped conceptual daily rainfall runoff model SIMHYD is shown in Figure 34 with its seven parameters highlighted in bold italics In SIMHYD daily rainfall first fills the interception store which is emptied each day by evaporation The excess rainfall is then subjected to an infiltration function that determines the infiltration capacity The excess rainfall that exceeds the infiltratio
120. ws users to configure RRL display options such as colours display fonts graph colours and graph fonts It also allows the user to specify the path to the default data directory looks Help Configuration i odel Data Path Figure 13 Tools drop down menu Help drop down menu The Help drop down menu is shown in Figure 14 This menu allows users to browse and search the RRL help file link to the Toolkit web site for information downloads and technical support and find out about the RRL The about screen lists the organisations and people that have contributed to the development of this product as well as a button that displays the licence agreement Content Search Index Toolkit Web Site gt Technical Support About RRL Figure 14 Help drop down menu 3 1 3 RRL Dialogues The RRL contains a single tabbed dialogue as shown in Figure 15 The tab dialogue allows the user to rapidly move between model options and parameters There are six dialogues that specify e Model The model type to be used e Input The daily time series inputs of rainfall evapotranspiration and flow e Dates Calibration and verification start warm up and end dates e Calibration Model parameters calibration method optimisation method and a graph display window 19 Rainfall Runoff Library User Guide Sensitivity Model parameter parameter limits and objective function Also includes a sensitivity graph display window Simulation

Rainfall Runoff Library User Guide

Contents

Download Pdf Manuals

Related Search

Related Contents